Skip to content

GitLab

"git@developer.sourcefind.cn:wangsen/mineru.git" did not exist on "39b46ea98092946ff6c437a144b6735b734186c6"
Commit 63719f08 authored by Anirudh Vegesana's avatar Anirudh Vegesana Committed by Jaeyoun Kim
Browse files

YOLO Family: Updated model (#9923) 



* Update YOLO model

* Fix some docstrings

* Fix docstrings

* Address some of Dr. Davis' changes

* Give descriptive names to the test cases

* Fix bugs

* Fix YOLO head imports

* docstring and variable name updates

* docstring and variable name updates

* docstring and variable name updates
Co-authored-by: default avatarvishnubanna <banna3vishnu@gmail.com>
Co-authored-by: default avatarVishnu Banna <43182884+vishnubanna@users.noreply.github.com>
parent 27be57eb
Hide whitespace changes
Inline Side-by-side
Showing with 2590 additions and 633 deletions
official/vision/beta/projects/yolo/configs/backbones.py
View file @ 63719f08
...@@ -24,11 +24,11 @@ from official.vision.beta.configs import backbones ...@@ -24,11 +24,11 @@ from official.vision.beta.configs import backbones
@dataclasses.dataclass @dataclasses.dataclass
class DarkNet(hyperparams.Config): class Darknet(hyperparams.Config):
"""DarkNet config.""" """Darknet config."""
model_id: str = "darknet53" model_id: str = "darknet53"
@dataclasses.dataclass @dataclasses.dataclass
class Backbone(backbones.Backbone): class Backbone(backbones.Backbone):
darknet: DarkNet = DarkNet() darknet: Darknet = Darknet()
official/vision/beta/projects/yolo/configs/darknet_classification.py
View file @ 63719f08
...@@ -32,7 +32,7 @@ class ImageClassificationModel(hyperparams.Config): ...@@ -32,7 +32,7 @@ class ImageClassificationModel(hyperparams.Config):
num_classes: int = 0 num_classes: int = 0
input_size: List[int] = dataclasses.field(default_factory=list) input_size: List[int] = dataclasses.field(default_factory=list)
backbone: backbones.Backbone = backbones.Backbone( backbone: backbones.Backbone = backbones.Backbone(
type='darknet', resnet=backbones.DarkNet()) type='darknet', darknet=backbones.Darknet())
dropout_rate: float = 0.0 dropout_rate: float = 0.0
norm_activation: common.NormActivation = common.NormActivation() norm_activation: common.NormActivation = common.NormActivation()
# Adds a BatchNormalization layer pre-GlobalAveragePooling in classification # Adds a BatchNormalization layer pre-GlobalAveragePooling in classification
......
official/vision/beta/projects/yolo/modeling/backbones/darknet.py
View file @ 63719f08
...@@ -29,44 +29,51 @@ Cross Stage Partial networks (CSPNets) were proposed in: ...@@ -29,44 +29,51 @@ Cross Stage Partial networks (CSPNets) were proposed in:
arXiv:1911.11929 arXiv:1911.11929
DarkNets Are used mainly for Object detection in: Darknets are used mainly for object detection in:
[1] Joseph Redmon, Ali Farhadi [1] Joseph Redmon, Ali Farhadi
YOLOv3: An Incremental Improvement. arXiv:1804.02767 YOLOv3: An Incremental Improvement. arXiv:1804.02767
[2] Alexey Bochkovskiy, Chien-Yao Wang, Hong-Yuan Mark Liao [2] Alexey Bochkovskiy, Chien-Yao Wang, Hong-Yuan Mark Liao
YOLOv4: Optimal Speed and Accuracy of Object Detection. arXiv:2004.10934 YOLOv4: Optimal Speed and Accuracy of Object Detection. arXiv:2004.10934
""" """
import collections
import collections
import tensorflow as tf import tensorflow as tf
from official.modeling import hyperparams from official.modeling import hyperparams
from official.vision.beta.modeling.backbones import factory from official.vision.beta.modeling.backbones import factory
from official.vision.beta.projects.yolo.modeling.layers import nn_blocks from official.vision.beta.projects.yolo.modeling.layers import nn_blocks
# builder required classes
class BlockConfig(object):
"""Get layer config to make code more readable.
Args: class BlockConfig:
layer: string layer name """
stack: the type of layer ordering to use for this specific level Class to store layer config to make code more readable
repetitions: integer for the number of times to repeat block
bottelneck: boolean for does this stack have a bottle neck layer
filters: integer for the output depth of the level
pool_size: integer the pool_size of max pool layers
kernel_size: optional integer, for convolution kernel size
strides: integer or tuple to indicate convolution strides
padding: the padding to apply to layers in this stack
activation: string for the activation to use for this stack
route: integer for what level to route from to get the next input
output_name: the name to use for this output
is_output: is this layer an output in the default model
""" """
def __init__(self, layer, stack, reps, bottleneck, filters, pool_size, def __init__(self, layer, stack, reps, bottleneck, filters, pool_size,
kernel_size, strides, padding, activation, route, output_name, kernel_size, strides, padding, activation, route, dilation_rate,
is_output): output_name, is_output):
"""
Args:
layer: A `str` for layer name.
stack: A `str` for the type of layer ordering to use for this specific
level.
reps: An `int` for the number of times to repeat block.
bottleneck: A `bool` for whether this stack has a bottle neck layer.
filters: An `int` for the output depth of the level.
pool_size: An `int` for the pool_size of max pool layers.
kernel_size: An `int` for convolution kernel size.
strides: A `Union[int, tuple]` that indicates convolution strides.
padding: An `int` for the padding to apply to layers in this stack.
activation: A `str` for the activation to use for this stack.
route: An `int` for the level to route from to get the next input.
dilation_rate: An `int` for the scale used in dialated Darknet.
output_name: A `str` for the name to use for this output.
is_output: A `bool` for whether this layer is an output in the default
model.
"""
self.layer = layer self.layer = layer
self.stack = stack self.stack = stack
self.repetitions = reps self.repetitions = reps
...@@ -78,6 +85,7 @@ class BlockConfig(object): ...@@ -78,6 +85,7 @@ class BlockConfig(object):
self.padding = padding self.padding = padding
self.activation = activation self.activation = activation
self.route = route self.route = route
self.dilation_rate = dilation_rate
self.output_name = output_name self.output_name = output_name
self.is_output = is_output self.is_output = is_output
...@@ -89,41 +97,40 @@ def build_block_specs(config): ...@@ -89,41 +97,40 @@ def build_block_specs(config):
return specs return specs
class LayerFactory(object): class LayerBuilder:
"""Class for quick look up of default layers. """
class for quick look up of default layers used by darknet to
Used by darknet to connect, introduce or exit a level. Used in place of an if connect, introduce or exit a level. Used in place of an if condition
condition or switch to make adding new layers easier and to reduce redundant or switch to make adding new layers easier and to reduce redundant code
code.
""" """
def __init__(self): def __init__(self):
self._layer_dict = { self._layer_dict = {
"ConvBN": (nn_blocks.ConvBN, self.conv_bn_config_todict), 'ConvBN': (nn_blocks.ConvBN, self.conv_bn_config_todict),
"MaxPool": (tf.keras.layers.MaxPool2D, self.maxpool_config_todict) 'MaxPool': (tf.keras.layers.MaxPool2D, self.maxpool_config_todict)
} }
def conv_bn_config_todict(self, config, kwargs): def conv_bn_config_todict(self, config, kwargs):
dictvals = { dictvals = {
"filters": config.filters, 'filters': config.filters,
"kernel_size": config.kernel_size, 'kernel_size': config.kernel_size,
"strides": config.strides, 'strides': config.strides,
"padding": config.padding 'padding': config.padding
} }
dictvals.update(kwargs) dictvals.update(kwargs)
return dictvals return dictvals
def darktiny_config_todict(self, config, kwargs): def darktiny_config_todict(self, config, kwargs):
dictvals = {"filters": config.filters, "strides": config.strides} dictvals = {'filters': config.filters, 'strides': config.strides}
dictvals.update(kwargs) dictvals.update(kwargs)
return dictvals return dictvals
def maxpool_config_todict(self, config, kwargs): def maxpool_config_todict(self, config, kwargs):
return { return {
"pool_size": config.pool_size, 'pool_size': config.pool_size,
"strides": config.strides, 'strides': config.strides,
"padding": config.padding, 'padding': config.padding,
"name": kwargs["name"] 'name': kwargs['name']
} }
def __call__(self, config, kwargs): def __call__(self, config, kwargs):
...@@ -134,90 +141,259 @@ class LayerFactory(object): ...@@ -134,90 +141,259 @@ class LayerFactory(object):
# model configs # model configs
LISTNAMES = [ LISTNAMES = [
"default_layer_name", "level_type", "number_of_layers_in_level", 'default_layer_name', 'level_type', 'number_of_layers_in_level',
"bottleneck", "filters", "kernal_size", "pool_size", "strides", "padding", 'bottleneck', 'filters', 'kernal_size', 'pool_size', 'strides', 'padding',
"default_activation", "route", "level/name", "is_output" 'default_activation', 'route', 'dilation', 'level/name', 'is_output'
] ]
# pylint: disable=line-too-long
CSPDARKNET53 = { CSPDARKNET53 = {
"list_names": LISTNAMES, 'list_names':
"splits": {"backbone_split": 106, LISTNAMES,
"neck_split": 138}, 'splits': {
"backbone": [ 'backbone_split': 106,
["ConvBN", None, 1, False, 32, None, 3, 1, "same", "mish", -1, 0, False], 'neck_split': 132
["DarkRes", "csp", 1, True, 64, None, None, None, None, "mish", -1, 1, False], },
["DarkRes", "csp", 2, False, 128, None, None, None, None, "mish", -1, 2, False], 'backbone': [
["DarkRes", "csp", 8, False, 256, None, None, None, None, "mish", -1, 3, True], [
["DarkRes", "csp", 8, False, 512, None, None, None, None, "mish", -1, 4, True], 'ConvBN', None, 1, False, 32, None, 3, 1, 'same', 'mish', -1, 1, 0,
["DarkRes", "csp", 4, False, 1024, None, None, None, None, "mish", -1, 5, True], False
],
[
'DarkRes', 'csp', 1, True, 64, None, None, None, None, 'mish', -1,
1, 1, False
],
[
'DarkRes', 'csp', 2, False, 128, None, None, None, None, 'mish', -1,
1, 2, False
],
[
'DarkRes', 'csp', 8, False, 256, None, None, None, None, 'mish', -1,
1, 3, True
],
[
'DarkRes', 'csp', 8, False, 512, None, None, None, None, 'mish', -1,
2, 4, True
],
[
'DarkRes', 'csp', 4, False, 1024, None, None, None, None, 'mish',
-1, 4, 5, True
],
]
}
CSPADARKNET53 = {
'list_names':
LISTNAMES,
'splits': {
'backbone_split': 100,
'neck_split': 135
},
'backbone': [
[
'ConvBN', None, 1, False, 32, None, 3, 1, 'same', 'mish', -1, 1, 0,
False
],
[
'DarkRes', 'residual', 1, True, 64, None, None, None, None, 'mish',
-1, 1, 1, False
],
[
'DarkRes', 'csp', 2, False, 128, None, None, None, None, 'mish', -1,
1, 2, False
],
[
'DarkRes', 'csp', 8, False, 256, None, None, None, None, 'mish', -1,
1, 3, True
],
[
'DarkRes', 'csp', 8, False, 512, None, None, None, None, 'mish', -1,
2, 4, True
],
[
'DarkRes', 'csp', 4, False, 1024, None, None, None, None, 'mish',
-1, 4, 5, True
],
]
}
LARGECSP53 = {
'list_names':
LISTNAMES,
'splits': {
'backbone_split': 100,
'neck_split': 135
},
'backbone': [
[
'ConvBN', None, 1, False, 32, None, 3, 1, 'same', 'mish', -1, 1, 0,
False
],
[
'DarkRes', 'csp', 1, True, 64, None, None, None, None, 'mish', -1,
1, 1, False
],
[
'DarkRes', 'csp', 3, False, 128, None, None, None, None, 'mish', -1,
1, 2, False
],
[
'DarkRes', 'csp', 15, False, 256, None, None, None, None, 'mish',
-1, 1, 3, True
],
[
'DarkRes', 'csp', 15, False, 512, None, None, None, None, 'mish',
-1, 2, 4, True
],
[
'DarkRes', 'csp', 7, False, 1024, None, None, None, None, 'mish',
-1, 4, 5, True
],
[
'DarkRes', 'csp', 7, False, 1024, None, None, None, None, 'mish',
-1, 8, 6, True
],
[
'DarkRes', 'csp', 7, False, 1024, None, None, None, None, 'mish',
-1, 16, 7, True
],
] ]
} }
DARKNET53 = { DARKNET53 = {
"list_names": LISTNAMES, 'list_names':
"splits": {"backbone_split": 76}, LISTNAMES,
"backbone": [ 'splits': {
["ConvBN", None, 1, False, 32, None, 3, 1, "same", "leaky", -1, 0, False], 'backbone_split': 76
["DarkRes", "residual", 1, True, 64, None, None, None, None, "leaky", -1, 1, False], },
["DarkRes", "residual", 2, False, 128, None, None, None, None, "leaky", -1, 2, False], 'backbone': [
["DarkRes", "residual", 8, False, 256, None, None, None, None, "leaky", -1, 3, True], [
["DarkRes", "residual", 8, False, 512, None, None, None, None, "leaky", -1, 4, True], 'ConvBN', None, 1, False, 32, None, 3, 1, 'same', 'leaky', -1, 1, 0,
["DarkRes", "residual", 4, False, 1024, None, None, None, None, "leaky", -1, 5, True], False
],
[
'DarkRes', 'residual', 1, True, 64, None, None, None, None, 'leaky',
-1, 1, 1, False
],
[
'DarkRes', 'residual', 2, False, 128, None, None, None, None,
'leaky', -1, 1, 2, False
],
[
'DarkRes', 'residual', 8, False, 256, None, None, None, None,
'leaky', -1, 1, 3, True
],
[
'DarkRes', 'residual', 8, False, 512, None, None, None, None,
'leaky', -1, 2, 4, True
],
[
'DarkRes', 'residual', 4, False, 1024, None, None, None, None,
'leaky', -1, 4, 5, True
],
] ]
} }
CSPDARKNETTINY = { CSPDARKNETTINY = {
"list_names": LISTNAMES, 'list_names':
"splits": {"backbone_split": 28}, LISTNAMES,
"backbone": [ 'splits': {
["ConvBN", None, 1, False, 32, None, 3, 2, "same", "leaky", -1, 0, False], 'backbone_split': 28
["ConvBN", None, 1, False, 64, None, 3, 2, "same", "leaky", -1, 1, False], },
["CSPTiny", "csp_tiny", 1, False, 64, None, 3, 2, "same", "leaky", -1, 2, False], 'backbone': [
["CSPTiny", "csp_tiny", 1, False, 128, None, 3, 2, "same", "leaky", -1, 3, False], [
["CSPTiny", "csp_tiny", 1, False, 256, None, 3, 2, "same", "leaky", -1, 4, True], 'ConvBN', None, 1, False, 32, None, 3, 2, 'same', 'leaky', -1, 1, 0,
["ConvBN", None, 1, False, 512, None, 3, 1, "same", "leaky", -1, 5, True], False
],
[
'ConvBN', None, 1, False, 64, None, 3, 2, 'same', 'leaky', -1, 1, 1,
False
],
[
'CSPTiny', 'csp_tiny', 1, False, 64, None, 3, 2, 'same', 'leaky',
-1, 1, 2, False
],
[
'CSPTiny', 'csp_tiny', 1, False, 128, None, 3, 2, 'same', 'leaky',
-1, 1, 3, False
],
[
'CSPTiny', 'csp_tiny', 1, False, 256, None, 3, 2, 'same', 'leaky',
-1, 1, 4, True
],
[
'ConvBN', None, 1, False, 512, None, 3, 1, 'same', 'leaky', -1, 1,
5, True
],
] ]
} }
DARKNETTINY = { DARKNETTINY = {
"list_names": LISTNAMES, 'list_names':
"splits": {"backbone_split": 14}, LISTNAMES,
"backbone": [ 'splits': {
["ConvBN", None, 1, False, 16, None, 3, 1, "same", "leaky", -1, 0, False], 'backbone_split': 14
["DarkTiny", "tiny", 1, True, 32, None, 3, 2, "same", "leaky", -1, 1, False], },
["DarkTiny", "tiny", 1, True, 64, None, 3, 2, "same", "leaky", -1, 2, False], 'backbone': [
["DarkTiny", "tiny", 1, False, 128, None, 3, 2, "same", "leaky", -1, 3, False], [
["DarkTiny", "tiny", 1, False, 256, None, 3, 2, "same", "leaky", -1, 4, True], 'ConvBN', None, 1, False, 16, None, 3, 1, 'same', 'leaky', -1, 1, 0,
["DarkTiny", "tiny", 1, False, 512, None, 3, 2, "same", "leaky", -1, 5, False], False
["DarkTiny", "tiny", 1, False, 1024, None, 3, 1, "same", "leaky", -1, 5, True], ],
[
'DarkTiny', 'tiny', 1, True, 32, None, 3, 2, 'same', 'leaky', -1, 1,
1, False
],
[
'DarkTiny', 'tiny', 1, True, 64, None, 3, 2, 'same', 'leaky', -1, 1,
2, False
],
[
'DarkTiny', 'tiny', 1, False, 128, None, 3, 2, 'same', 'leaky', -1,
1, 3, False
],
[
'DarkTiny', 'tiny', 1, False, 256, None, 3, 2, 'same', 'leaky', -1,
1, 4, True
],
[
'DarkTiny', 'tiny', 1, False, 512, None, 3, 2, 'same', 'leaky', -1,
1, 5, False
],
[
'DarkTiny', 'tiny', 1, False, 1024, None, 3, 1, 'same', 'leaky', -1,
1, 5, True
],
] ]
} }
# pylint: enable=line-too-long
BACKBONES = { BACKBONES = {
"darknettiny": DARKNETTINY, 'darknettiny': DARKNETTINY,
"darknet53": DARKNET53, 'darknet53': DARKNET53,
"cspdarknet53": CSPDARKNET53, 'cspdarknet53': CSPDARKNET53,
"cspdarknettiny": CSPDARKNETTINY 'altered_cspdarknet53': CSPADARKNET53,
'cspdarknettiny': CSPDARKNETTINY,
'csp-large': LARGECSP53,
} }
@tf.keras.utils.register_keras_serializable(package="yolo") @tf.keras.utils.register_keras_serializable(package='yolo')
class Darknet(tf.keras.Model): class Darknet(tf.keras.Model):
"""Darknet backbone.""" """ The Darknet backbone architecture """
def __init__( def __init__(
self, self,
model_id="darknet53", model_id='darknet53',
input_specs=tf.keras.layers.InputSpec(shape=[None, None, None, 3]), input_specs=tf.keras.layers.InputSpec(shape=[None, None, None, 3]),
min_level=None, min_level=None,
max_level=5, max_level=5,
width_scale=1.0,
depth_scale=1.0,
csp_level_mod=(),
activation=None, activation=None,
use_sync_bn=False, use_sync_bn=False,
norm_momentum=0.99, norm_momentum=0.99,
norm_epsilon=0.001, norm_epsilon=0.001,
kernel_initializer="glorot_uniform", dilate=False,
kernel_initializer='glorot_uniform',
kernel_regularizer=None, kernel_regularizer=None,
bias_regularizer=None, bias_regularizer=None,
**kwargs): **kwargs):
...@@ -227,12 +403,13 @@ class Darknet(tf.keras.Model): ...@@ -227,12 +403,13 @@ class Darknet(tf.keras.Model):
self._model_name = model_id self._model_name = model_id
self._splits = splits self._splits = splits
self._input_shape = input_specs self._input_shape = input_specs
self._registry = LayerFactory() self._registry = LayerBuilder()
# default layer look up # default layer look up
self._min_size = min_level self._min_size = min_level
self._max_size = max_level self._max_size = max_level
self._output_specs = None self._output_specs = None
self._csp_level_mod = set(csp_level_mod)
self._kernel_initializer = kernel_initializer self._kernel_initializer = kernel_initializer
self._bias_regularizer = bias_regularizer self._bias_regularizer = bias_regularizer
...@@ -241,16 +418,20 @@ class Darknet(tf.keras.Model): ...@@ -241,16 +418,20 @@ class Darknet(tf.keras.Model):
self._use_sync_bn = use_sync_bn self._use_sync_bn = use_sync_bn
self._activation = activation self._activation = activation
self._kernel_regularizer = kernel_regularizer self._kernel_regularizer = kernel_regularizer
self._dilate = dilate
self._width_scale = width_scale
self._depth_scale = depth_scale
self._default_dict = { self._default_dict = {
"kernel_initializer": self._kernel_initializer, 'kernel_initializer': self._kernel_initializer,
"kernel_regularizer": self._kernel_regularizer, 'kernel_regularizer': self._kernel_regularizer,
"bias_regularizer": self._bias_regularizer, 'bias_regularizer': self._bias_regularizer,
"norm_momentum": self._norm_momentum, 'norm_momentum': self._norm_momentum,
"norm_epsilon": self._norm_epislon, 'norm_epsilon': self._norm_epislon,
"use_sync_bn": self._use_sync_bn, 'use_sync_bn': self._use_sync_bn,
"activation": self._activation, 'activation': self._activation,
"name": None 'dilation_rate': 1,
'name': None
} }
inputs = tf.keras.layers.Input(shape=self._input_shape.shape[1:]) inputs = tf.keras.layers.Input(shape=self._input_shape.shape[1:])
...@@ -273,33 +454,39 @@ class Darknet(tf.keras.Model): ...@@ -273,33 +454,39 @@ class Darknet(tf.keras.Model):
endpoints = collections.OrderedDict() endpoints = collections.OrderedDict()
stack_outputs = [inputs] stack_outputs = [inputs]
for i, config in enumerate(net): for i, config in enumerate(net):
if config.output_name > self._max_size:
break
if config.output_name in self._csp_level_mod:
config.stack = 'residual'
config.filters = int(config.filters * self._width_scale)
config.repetitions = int(config.repetitions * self._depth_scale)
if config.stack is None: if config.stack is None:
x = self._build_block(stack_outputs[config.route], x = self._build_block(
config, stack_outputs[config.route], config, name=f'{config.layer}_{i}')
name=f"{config.layer}_{i}")
stack_outputs.append(x) stack_outputs.append(x)
elif config.stack == "residual": elif config.stack == 'residual':
x = self._residual_stack(stack_outputs[config.route], x = self._residual_stack(
config, stack_outputs[config.route], config, name=f'{config.layer}_{i}')
name=f"{config.layer}_{i}")
stack_outputs.append(x) stack_outputs.append(x)
elif config.stack == "csp": elif config.stack == 'csp':
x = self._csp_stack(stack_outputs[config.route], x = self._csp_stack(
config, stack_outputs[config.route], config, name=f'{config.layer}_{i}')
name=f"{config.layer}_{i}")
stack_outputs.append(x) stack_outputs.append(x)
elif config.stack == "csp_tiny": elif config.stack == 'csp_tiny':
x_pass, x = self._csp_tiny_stack(stack_outputs[config.route], x_pass, x = self._csp_tiny_stack(
config, name=f"{config.layer}_{i}") stack_outputs[config.route], config, name=f'{config.layer}_{i}')
stack_outputs.append(x_pass) stack_outputs.append(x_pass)
elif config.stack == "tiny": elif config.stack == 'tiny':
x = self._tiny_stack(stack_outputs[config.route], x = self._tiny_stack(
config, stack_outputs[config.route], config, name=f'{config.layer}_{i}')
name=f"{config.layer}_{i}")
stack_outputs.append(x) stack_outputs.append(x)
if (config.is_output and self._min_size is None): if (config.is_output and self._min_size is None):
endpoints[str(config.output_name)] = x endpoints[str(config.output_name)] = x
elif self._min_size is not None and config.output_name >= self._min_size and config.output_name <= self._max_size: elif (self._min_size is not None and
config.output_name >= self._min_size and
config.output_name <= self._max_size):
endpoints[str(config.output_name)] = x endpoints[str(config.output_name)] = x
self._output_specs = {l: endpoints[l].get_shape() for l in endpoints.keys()} self._output_specs = {l: endpoints[l].get_shape() for l in endpoints.keys()}
...@@ -308,8 +495,7 @@ class Darknet(tf.keras.Model): ...@@ -308,8 +495,7 @@ class Darknet(tf.keras.Model):
def _get_activation(self, activation): def _get_activation(self, activation):
if self._activation is None: if self._activation is None:
return activation return activation
else: return self._activation
return self._activation
def _csp_stack(self, inputs, config, name): def _csp_stack(self, inputs, config, name):
if config.bottleneck: if config.bottleneck:
...@@ -320,86 +506,135 @@ class Darknet(tf.keras.Model): ...@@ -320,86 +506,135 @@ class Darknet(tf.keras.Model):
csp_filter_scale = 2 csp_filter_scale = 2
residual_filter_scale = 1 residual_filter_scale = 1
scale_filters = 2 scale_filters = 2
self._default_dict["activation"] = self._get_activation(config.activation) self._default_dict['activation'] = self._get_activation(config.activation)
self._default_dict["name"] = f"{name}_csp_down" self._default_dict['name'] = f'{name}_csp_down'
x, x_route = nn_blocks.CSPRoute(filters=config.filters, if self._dilate:
filter_scale=csp_filter_scale, self._default_dict['dilation_rate'] = config.dilation_rate
downsample=True, else:
**self._default_dict)(inputs) self._default_dict['dilation_rate'] = 1
for i in range(config.repetitions):
self._default_dict["name"] = f"{name}_{i}" # swap/add dilation
x = nn_blocks.DarkResidual(filters=config.filters // scale_filters, x, x_route = nn_blocks.CSPRoute(
filter_scale=residual_filter_scale, filters=config.filters,
**self._default_dict)(x) filter_scale=csp_filter_scale,
downsample=True,
self._default_dict["name"] = f"{name}_csp_connect" **self._default_dict)(
output = nn_blocks.CSPConnect(filters=config.filters, inputs)
filter_scale=csp_filter_scale,
**self._default_dict)([x, x_route]) dilated_reps = config.repetitions - self._default_dict['dilation_rate'] // 2
self._default_dict["activation"] = self._activation for i in range(dilated_reps):
self._default_dict["name"] = None self._default_dict['name'] = f'{name}_{i}'
x = nn_blocks.DarkResidual(
filters=config.filters // scale_filters,
filter_scale=residual_filter_scale,
**self._default_dict)(
x)
for i in range(dilated_reps, config.repetitions):
self._default_dict[
'dilation_rate'] = self._default_dict['dilation_rate'] // 2
self._default_dict[
'name'] = f"{name}_{i}_degridded_{self._default_dict['dilation_rate']}"
x = nn_blocks.DarkResidual(
filters=config.filters // scale_filters,
filter_scale=residual_filter_scale,
**self._default_dict)(
x)
self._default_dict['name'] = f'{name}_csp_connect'
output = nn_blocks.CSPConnect(
filters=config.filters,
filter_scale=csp_filter_scale,
**self._default_dict)([x, x_route])
self._default_dict['activation'] = self._activation
self._default_dict['name'] = None
return output return output
def _csp_tiny_stack(self, inputs, config, name): def _csp_tiny_stack(self, inputs, config, name):
self._default_dict["activation"] = self._get_activation(config.activation) self._default_dict['activation'] = self._get_activation(config.activation)
self._default_dict["name"] = f"{name}_csp_tiny" self._default_dict['name'] = f'{name}_csp_tiny'
x, x_route = nn_blocks.CSPTiny(filters=config.filters, x, x_route = nn_blocks.CSPTiny(
**self._default_dict)(inputs) filters=config.filters, **self._default_dict)(
self._default_dict["activation"] = self._activation inputs)
self._default_dict["name"] = None self._default_dict['activation'] = self._activation
self._default_dict['name'] = None
return x, x_route return x, x_route
def _tiny_stack(self, inputs, config, name): def _tiny_stack(self, inputs, config, name):
x = tf.keras.layers.MaxPool2D(pool_size=2, x = tf.keras.layers.MaxPool2D(
strides=config.strides, pool_size=2,
padding="same", strides=config.strides,
data_format=None, padding='same',
name=f"{name}_tiny/pool")(inputs) data_format=None,
self._default_dict["activation"] = self._get_activation(config.activation) name=f'{name}_tiny/pool')(
self._default_dict["name"] = f"{name}_tiny/conv" inputs)
self._default_dict['activation'] = self._get_activation(config.activation)
self._default_dict['name'] = f'{name}_tiny/conv'
x = nn_blocks.ConvBN( x = nn_blocks.ConvBN(
filters=config.filters, filters=config.filters,
kernel_size=(3, 3), kernel_size=(3, 3),
strides=(1, 1), strides=(1, 1),
padding="same", padding='same',
**self._default_dict)( **self._default_dict)(
x) x)
self._default_dict["activation"] = self._activation self._default_dict['activation'] = self._activation
self._default_dict["name"] = None self._default_dict['name'] = None
return x return x
def _residual_stack(self, inputs, config, name): def _residual_stack(self, inputs, config, name):
self._default_dict["activation"] = self._get_activation(config.activation) self._default_dict['activation'] = self._get_activation(config.activation)
self._default_dict["name"] = f"{name}_residual_down" self._default_dict['name'] = f'{name}_residual_down'
x = nn_blocks.DarkResidual(filters=config.filters, if self._dilate:
downsample=True, self._default_dict['dilation_rate'] = config.dilation_rate
**self._default_dict)(inputs) if config.repetitions < 8:
for i in range(config.repetitions - 1): config.repetitions += 2
self._default_dict["name"] = f"{name}_{i}" else:
x = nn_blocks.DarkResidual(filters=config.filters, self._default_dict['dilation_rate'] = 1
**self._default_dict)(x)
self._default_dict["activation"] = self._activation x = nn_blocks.DarkResidual(
self._default_dict["name"] = None filters=config.filters, downsample=True, **self._default_dict)(
inputs)
dilated_reps = config.repetitions - \
(self._default_dict['dilation_rate'] // 2) - 1
for i in range(dilated_reps):
self._default_dict['name'] = f'{name}_{i}'
x = nn_blocks.DarkResidual(
filters=config.filters, **self._default_dict)(
x)
for i in range(dilated_reps, config.repetitions - 1):
self._default_dict[
'dilation_rate'] = self._default_dict['dilation_rate'] // 2
self._default_dict[
'name'] = f"{name}_{i}_degridded_{self._default_dict['dilation_rate']}"
x = nn_blocks.DarkResidual(
filters=config.filters, **self._default_dict)(
x)
self._default_dict['activation'] = self._activation
self._default_dict['name'] = None
self._default_dict['dilation_rate'] = 1
return x return x
def _build_block(self, inputs, config, name): def _build_block(self, inputs, config, name):
x = inputs x = inputs
i = 0 i = 0
self._default_dict["activation"] = self._get_activation(config.activation) self._default_dict['activation'] = self._get_activation(config.activation)
while i < config.repetitions: while i < config.repetitions:
self._default_dict["name"] = f"{name}_{i}" self._default_dict['name'] = f'{name}_{i}'
layer = self._registry(config, self._default_dict) layer = self._registry(config, self._default_dict)
x = layer(x) x = layer(x)
i += 1 i += 1
self._default_dict["activation"] = self._activation self._default_dict['activation'] = self._activation
self._default_dict["name"] = None self._default_dict['name'] = None
return x return x
@staticmethod @staticmethod
def get_model_config(name): def get_model_config(name):
name = name.lower() name = name.lower()
backbone = BACKBONES[name]["backbone"] backbone = BACKBONES[name]['backbone']
splits = BACKBONES[name]["splits"] splits = BACKBONES[name]['splits']
return build_block_specs(backbone), splits return build_block_specs(backbone), splits
@property @property
...@@ -412,35 +647,42 @@ class Darknet(tf.keras.Model): ...@@ -412,35 +647,42 @@ class Darknet(tf.keras.Model):
def get_config(self): def get_config(self):
layer_config = { layer_config = {
"model_id": self._model_name, 'model_id': self._model_name,
"min_level": self._min_size, 'min_level': self._min_size,
"max_level": self._max_size, 'max_level': self._max_size,
"kernel_initializer": self._kernel_initializer, 'kernel_initializer': self._kernel_initializer,
"kernel_regularizer": self._kernel_regularizer, 'kernel_regularizer': self._kernel_regularizer,
"bias_regularizer": self._bias_regularizer, 'bias_regularizer': self._bias_regularizer,
"norm_momentum": self._norm_momentum, 'norm_momentum': self._norm_momentum,
"norm_epsilon": self._norm_epislon, 'norm_epsilon': self._norm_epislon,
"use_sync_bn": self._use_sync_bn, 'use_sync_bn': self._use_sync_bn,
"activation": self._activation 'activation': self._activation,
} }
return layer_config return layer_config
@factory.register_backbone_builder("darknet") @factory.register_backbone_builder('darknet')
def build_darknet( def build_darknet(
input_specs: tf.keras.layers.InputSpec, input_specs: tf.keras.layers.InputSpec,
backbone_config: hyperparams.Config, backbone_config: hyperparams.Config,
norm_activation_config: hyperparams.Config, norm_activation_config: hyperparams.Config,
l2_regularizer: tf.keras.regularizers.Regularizer = None) -> tf.keras.Model: l2_regularizer: tf.keras.regularizers.Regularizer = None) -> tf.keras.Model:
"""Builds darknet backbone."""
backbone_cfg = backbone_config.get() backbone_cfg = model_config.backbone.get()
norm_activation_config = model_config.norm_activation
model = Darknet( model = Darknet(
model_id=backbone_cfg.model_id, model_id=backbone_cfg.model_id,
input_shape=input_specs, min_level=model_config.min_level,
max_level=model_config.max_level,
input_specs=input_specs,
dilate=backbone_cfg.dilate,
width_scale=backbone_cfg.width_scale,
depth_scale=backbone_cfg.depth_scale,
activation=norm_activation_config.activation, activation=norm_activation_config.activation,
use_sync_bn=norm_activation_config.use_sync_bn, use_sync_bn=norm_activation_config.use_sync_bn,
norm_momentum=norm_activation_config.norm_momentum, norm_momentum=norm_activation_config.norm_momentum,
norm_epsilon=norm_activation_config.norm_epsilon, norm_epsilon=norm_activation_config.norm_epsilon,
kernel_regularizer=l2_regularizer) kernel_regularizer=l2_regularizer)
model.summary()
return model return model
official/vision/beta/projects/yolo/modeling/backbones/darknet_test.py
View file @ 63719f08
...@@ -13,7 +13,7 @@ ...@@ -13,7 +13,7 @@
# limitations under the License. # limitations under the License.
# Lint as: python3 # Lint as: python3
"""Tests for resnet.""" """Tests for yolo."""
from absl.testing import parameterized from absl.testing import parameterized
import numpy as np import numpy as np
...@@ -24,40 +24,53 @@ from tensorflow.python.distribute import strategy_combinations ...@@ -24,40 +24,53 @@ from tensorflow.python.distribute import strategy_combinations
from official.vision.beta.projects.yolo.modeling.backbones import darknet from official.vision.beta.projects.yolo.modeling.backbones import darknet
class DarkNetTest(parameterized.TestCase, tf.test.TestCase): class DarknetTest(parameterized.TestCase, tf.test.TestCase):
@parameterized.parameters( @parameterized.parameters(
(224, "darknet53", 2, 1), (224, 'darknet53', 2, 1, True),
(224, "darknettiny", 1, 2), (224, 'darknettiny', 1, 2, False),
(224, "cspdarknettiny", 1, 1), (224, 'cspdarknettiny', 1, 1, False),
(224, "cspdarknet53", 2, 1), (224, 'cspdarknet53', 2, 1, True),
) )
def test_network_creation(self, input_size, model_id, def test_network_creation(self, input_size, model_id, endpoint_filter_scale,
endpoint_filter_scale, scale_final): scale_final, dilate):
"""Test creation of ResNet family models.""" """Test creation of ResNet family models."""
tf.keras.backend.set_image_data_format("channels_last") tf.keras.backend.set_image_data_format('channels_last')
network = darknet.Darknet(model_id=model_id, min_level=3, max_level=5) network = darknet.Darknet(
model_id=model_id, min_level=3, max_level=5, dilate=dilate)
self.assertEqual(network.model_id, model_id) self.assertEqual(network.model_id, model_id)
inputs = tf.keras.Input(shape=(input_size, input_size, 3), batch_size=1) inputs = tf.keras.Input(shape=(input_size, input_size, 3), batch_size=1)
endpoints = network(inputs) endpoints = network(inputs)
self.assertAllEqual( if dilate:
[1, input_size / 2**3, input_size / 2**3, 128 * endpoint_filter_scale], self.assertAllEqual([
endpoints["3"].shape.as_list()) 1, input_size / 2**3, input_size / 2**3, 128 * endpoint_filter_scale
self.assertAllEqual( ], endpoints['3'].shape.as_list())
[1, input_size / 2**4, input_size / 2**4, 256 * endpoint_filter_scale], self.assertAllEqual([
endpoints["4"].shape.as_list()) 1, input_size / 2**3, input_size / 2**3, 256 * endpoint_filter_scale
self.assertAllEqual([ ], endpoints['4'].shape.as_list())
1, input_size / 2**5, input_size / 2**5, self.assertAllEqual([
512 * endpoint_filter_scale * scale_final 1, input_size / 2**3, input_size / 2**3,
], endpoints["5"].shape.as_list()) 512 * endpoint_filter_scale * scale_final
], endpoints['5'].shape.as_list())
else:
self.assertAllEqual([
1, input_size / 2**3, input_size / 2**3, 128 * endpoint_filter_scale
], endpoints['3'].shape.as_list())
self.assertAllEqual([
1, input_size / 2**4, input_size / 2**4, 256 * endpoint_filter_scale
], endpoints['4'].shape.as_list())
self.assertAllEqual([
1, input_size / 2**5, input_size / 2**5,
512 * endpoint_filter_scale * scale_final
], endpoints['5'].shape.as_list())
@combinations.generate( @combinations.generate(
combinations.combine( combinations.combine(
strategy=[ strategy=[
strategy_combinations.cloud_tpu_strategy, strategy_combinations.tpu_strategy,
strategy_combinations.one_device_strategy_gpu, strategy_combinations.one_device_strategy_gpu,
], ],
use_sync_bn=[False, True], use_sync_bn=[False, True],
...@@ -66,20 +79,20 @@ class DarkNetTest(parameterized.TestCase, tf.test.TestCase): ...@@ -66,20 +79,20 @@ class DarkNetTest(parameterized.TestCase, tf.test.TestCase):
"""Test for sync bn on TPU and GPU devices.""" """Test for sync bn on TPU and GPU devices."""
inputs = np.random.rand(1, 224, 224, 3) inputs = np.random.rand(1, 224, 224, 3)
tf.keras.backend.set_image_data_format("channels_last") tf.keras.backend.set_image_data_format('channels_last')
with strategy.scope(): with strategy.scope():
network = darknet.Darknet(model_id="darknet53", min_size=3, max_size=5) network = darknet.Darknet(model_id='darknet53', min_size=3, max_size=5)
_ = network(inputs) _ = network(inputs)
@parameterized.parameters(1, 3, 4) @parameterized.parameters(1, 3, 4)
def test_input_specs(self, input_dim): def test_input_specs(self, input_dim):
"""Test different input feature dimensions.""" """Test different input feature dimensions."""
tf.keras.backend.set_image_data_format("channels_last") tf.keras.backend.set_image_data_format('channels_last')
input_specs = tf.keras.layers.InputSpec(shape=[None, None, None, input_dim]) input_specs = tf.keras.layers.InputSpec(shape=[None, None, None, input_dim])
network = darknet.Darknet( network = darknet.Darknet(
model_id="darknet53", min_level=3, max_level=5, input_specs=input_specs) model_id='darknet53', min_level=3, max_level=5, input_specs=input_specs)
inputs = tf.keras.Input(shape=(224, 224, input_dim), batch_size=1) inputs = tf.keras.Input(shape=(224, 224, input_dim), batch_size=1)
_ = network(inputs) _ = network(inputs)
...@@ -87,14 +100,14 @@ class DarkNetTest(parameterized.TestCase, tf.test.TestCase): ...@@ -87,14 +100,14 @@ class DarkNetTest(parameterized.TestCase, tf.test.TestCase):
def test_serialize_deserialize(self): def test_serialize_deserialize(self):
# Create a network object that sets all of its config options. # Create a network object that sets all of its config options.
kwargs = dict( kwargs = dict(
model_id="darknet53", model_id='darknet53',
min_level=3, min_level=3,
max_level=5, max_level=5,
use_sync_bn=False, use_sync_bn=False,
activation="relu", activation='relu',
norm_momentum=0.99, norm_momentum=0.99,
norm_epsilon=0.001, norm_epsilon=0.001,
kernel_initializer="VarianceScaling", kernel_initializer='VarianceScaling',
kernel_regularizer=None, kernel_regularizer=None,
bias_regularizer=None, bias_regularizer=None,
) )
...@@ -113,5 +126,5 @@ class DarkNetTest(parameterized.TestCase, tf.test.TestCase): ...@@ -113,5 +126,5 @@ class DarkNetTest(parameterized.TestCase, tf.test.TestCase):
self.assertAllEqual(network.get_config(), new_network.get_config()) self.assertAllEqual(network.get_config(), new_network.get_config())
if __name__ == "__main__": if __name__ == '__main__':
tf.test.main() tf.test.main()
official/vision/beta/projects/yolo/modeling/decoders/yolo_decoder.py 0 → 100644
View file @ 63719f08
# Copyright 2021 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# Lint as: python3
"""Feature Pyramid Network and Path Aggregation variants used in YOLO"""
import tensorflow as tf
from official.vision.beta.projects.yolo.modeling.layers import nn_blocks
@tf.keras.utils.register_keras_serializable(package='yolo')
class _IdentityRoute(tf.keras.layers.Layer):
def __init__(self, **kwargs):
"""Private class to mirror the outputs of blocks in nn_blocks for an easier
programatic generation of the feature pyramid network"""
super().__init__(**kwargs)
def call(self, inputs): # pylint: disable=arguments-differ
return None, inputs
@tf.keras.utils.register_keras_serializable(package='yolo')
class YoloFPN(tf.keras.layers.Layer):
"""YOLO Feature pyramid network."""
def __init__(self,
fpn_depth=4,
use_spatial_attention=False,
csp_stack=False,
activation='leaky',
fpn_filter_scale=1,
use_sync_bn=False,
norm_momentum=0.99,
norm_epsilon=0.001,
kernel_initializer='glorot_uniform',
kernel_regularizer=None,
bias_regularizer=None,
**kwargs):
"""Yolo FPN initialization function (Yolo V4).
Args:
fpn_depth: `int`, number of layers to use in each FPN path
if you choose to use an FPN.
use_spatial_attention: `bool`, use the spatial attention module.
csp_stack: `bool`, CSPize the FPN.
activation: `str`, the activation function to use typically leaky or mish.
fpn_filter_scale: `int`, scaling factor for the FPN filters.
use_sync_bn: if True, use synchronized batch normalization.
norm_momentum: `float`, normalization momentum for the moving average.
norm_epsilon: `float`, small float added to variance to avoid dividing by
zero.
kernel_initializer: kernel_initializer for convolutional layers.
kernel_regularizer: tf.keras.regularizers.Regularizer object for Conv2D.
bias_regularizer: tf.keras.regularizers.Regularizer object for Conv2D.
**kwargs: keyword arguments to be passed.
"""
super().__init__(**kwargs)
self._fpn_depth = fpn_depth
self._activation = activation
self._use_sync_bn = use_sync_bn
self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon
self._kernel_initializer = kernel_initializer
self._kernel_regularizer = kernel_regularizer
self._bias_regularizer = bias_regularizer
self._use_spatial_attention = use_spatial_attention
self._filter_scale = fpn_filter_scale
self._csp_stack = csp_stack
self._base_config = dict(
activation=self._activation,
use_sync_bn=self._use_sync_bn,
kernel_regularizer=self._kernel_regularizer,
kernel_initializer=self._kernel_initializer,
bias_regularizer=self._bias_regularizer,
norm_epsilon=self._norm_epsilon,
norm_momentum=self._norm_momentum)
def get_raw_depths(self, minimum_depth, inputs):
"""Calculates the unscaled depths of the FPN branches.
Args:
minimum_depth (int): depth of the smallest branch of the FPN.
inputs (dict): dictionary of the shape of input args as a dictionary of
lists.
Returns:
The unscaled depths of the FPN branches.
"""
depths = []
for i in range(self._min_level, self._max_level + 1):
depths.append(inputs[str(i)][-1] / self._filter_scale)
return list(reversed(depths))
def build(self, inputs):
"""Use config dictionary to generate all important attributes for head
construction.
Args:
inputs: dictionary of the shape of input args as a dictionary of lists.
"""
keys = [int(key) for key in inputs.keys()]
self._min_level = min(keys)
self._max_level = max(keys)
self._min_depth = inputs[str(self._min_level)][-1]
self._depths = self.get_raw_depths(self._min_depth, inputs)
# directly connect to an input path and process it
self.preprocessors = dict()
# resample an input and merge it with the output of another path
# inorder to aggregate backbone outputs
self.resamples = dict()
# set of convoltion layers and upsample layers that are used to
# prepare the FPN processors for output
for level, depth in zip(
reversed(range(self._min_level, self._max_level + 1)), self._depths):
if level == self._min_level:
self.resamples[str(level)] = nn_blocks.PathAggregationBlock(
filters=depth // 2,
inverted=True,
upsample=True,
drop_final=self._csp_stack == 0,
upsample_size=2,
**self._base_config)
self.preprocessors[str(level)] = _IdentityRoute()
elif level != self._max_level:
self.resamples[str(level)] = nn_blocks.PathAggregationBlock(
filters=depth // 2,
inverted=True,
upsample=True,
drop_final=False,
upsample_size=2,
**self._base_config)
self.preprocessors[str(level)] = nn_blocks.DarkRouteProcess(
filters=depth,
repetitions=self._fpn_depth - int(level == self._min_level),
block_invert=True,
insert_spp=False,
csp_stack=self._csp_stack,
**self._base_config)
else:
self.preprocessors[str(level)] = nn_blocks.DarkRouteProcess(
filters=depth,
repetitions=self._fpn_depth + 1 * int(self._csp_stack == 0),
insert_spp=True,
block_invert=False,
csp_stack=self._csp_stack,
**self._base_config)
def call(self, inputs):
outputs = dict()
layer_in = inputs[str(self._max_level)]
for level in reversed(range(self._min_level, self._max_level + 1)):
_, x = self.preprocessors[str(level)](layer_in)
outputs[str(level)] = x
if level > self._min_level:
x_next = inputs[str(level - 1)]
_, layer_in = self.resamples[str(level - 1)]([x_next, x])
return outputs
@tf.keras.utils.register_keras_serializable(package='yolo')
class YoloPAN(tf.keras.layers.Layer):
"""YOLO Path Aggregation Network"""
def __init__(self,
path_process_len=6,
max_level_process_len=None,
embed_spp=False,
use_spatial_attention=False,
csp_stack=False,
activation='leaky',
use_sync_bn=False,
norm_momentum=0.99,
norm_epsilon=0.001,
kernel_initializer='glorot_uniform',
kernel_regularizer=None,
bias_regularizer=None,
fpn_input=True,
fpn_filter_scale=1.0,
**kwargs):
"""Yolo Path Aggregation Network initialization function (Yolo V3 and V4).
Args:
path_process_len: `int`, number of layers ot use in each Decoder path.
max_level_process_len: `int`, number of layers ot use in the largest
processing path, or the backbones largest output if it is different.
embed_spp: `bool`, use the SPP found in the YoloV3 and V4 model.
use_spatial_attention: `bool`, use the spatial attention module.
csp_stack: `bool`, CSPize the FPN.
activation: `str`, the activation function to use typically leaky or mish.
use_sync_bn: if True, use synchronized batch normalization.
norm_momentum: `float`, normalization omentum for the moving average.
norm_epsilon: `float`, small float added to variance to avoid dividing
by zero.
kernel_initializer: kernel_initializer for convolutional layers.
kernel_regularizer: tf.keras.regularizers.Regularizer object for Conv2D.
bias_regularizer: tf.keras.regularizers.Regularizer object for Conv2d.
fpn_input: `bool`, for whether the input into this fucntion is an FPN or
a backbone.
fpn_filter_scale: `int`, scaling factor for the FPN filters.
**kwargs: keyword arguments to be passed.
"""
super().__init__(**kwargs)
self._path_process_len = path_process_len
self._embed_spp = embed_spp
self._use_spatial_attention = use_spatial_attention
self._activation = activation
self._use_sync_bn = use_sync_bn
self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon
self._kernel_initializer = kernel_initializer
self._kernel_regularizer = kernel_regularizer
self._bias_regularizer = bias_regularizer
self._fpn_input = fpn_input
self._max_level_process_len = max_level_process_len
self._csp_stack = csp_stack
self._fpn_filter_scale = fpn_filter_scale
if max_level_process_len is None:
self._max_level_process_len = path_process_len
self._base_config = dict(
activation=self._activation,
use_sync_bn=self._use_sync_bn,
kernel_regularizer=self._kernel_regularizer,
kernel_initializer=self._kernel_initializer,
bias_regularizer=self._bias_regularizer,
norm_epsilon=self._norm_epsilon,
norm_momentum=self._norm_momentum)
def build(self, inputs):
"""Use config dictionary to generate all important attributes for head
construction.
Args:
inputs: dictionary of the shape of input args as a dictionary of lists.
"""
# define the key order
keys = [int(key) for key in inputs.keys()]
self._min_level = min(keys)
self._max_level = max(keys)
self._min_depth = inputs[str(self._min_level)][-1]
self._depths = self.get_raw_depths(self._min_depth, inputs)
# directly connect to an input path and process it
self.preprocessors = dict()
# resample an input and merge it with the output of another path
# inorder to aggregate backbone outputs
self.resamples = dict()
# FPN will reverse the key process order for the backbone, so we need
# adjust the order that objects are created and processed to adjust for
# this. not using an FPN will directly connect the decoder to the backbone
# therefore the object creation order needs to be done from the largest
# to smallest level.
if self._fpn_input:
# process order {... 3, 4, 5}
self._iterator = range(self._min_level, self._max_level + 1)
self._check = lambda x: x < self._max_level
self._key_shift = lambda x: x + 1
self._input = self._min_level
downsample = True
upsample = False
else:
# process order {5, 4, 3, ...}
self._iterator = list(
reversed(range(self._min_level, self._max_level + 1)))
self._check = lambda x: x > self._min_level
self._key_shift = lambda x: x - 1
self._input = self._max_level
downsample = False
upsample = True
if self._csp_stack == 0:
proc_filters = lambda x: x
resample_filters = lambda x: x // 2
else:
proc_filters = lambda x: x * 2
resample_filters = lambda x: x
for level, depth in zip(self._iterator, self._depths):
if level == self._input:
self.preprocessors[str(level)] = nn_blocks.DarkRouteProcess(
filters=proc_filters(depth),
repetitions=self._max_level_process_len,
insert_spp=self._embed_spp,
block_invert=False,
insert_sam=self._use_spatial_attention,
csp_stack=self._csp_stack,
**self._base_config)
else:
self.resamples[str(level)] = nn_blocks.PathAggregationBlock(
filters=resample_filters(depth),
upsample=upsample,
downsample=downsample,
inverted=False,
drop_final=self._csp_stack == 0,
**self._base_config)
self.preprocessors[str(level)] = nn_blocks.DarkRouteProcess(
filters=proc_filters(depth),
repetitions=self._path_process_len,
insert_spp=False,
insert_sam=self._use_spatial_attention,
csp_stack=self._csp_stack,
**self._base_config)
def get_raw_depths(self, minimum_depth, inputs):
"""Calculates the unscaled depths of the FPN branches.
Args:
minimum_depth: `int` depth of the smallest branch of the FPN.
inputs: `dict[str, tf.InputSpec]` of the shape of input args as a dictionary of
lists.
Returns:
The unscaled depths of the FPN branches.
"""
depths = []
if len(inputs.keys()) > 3 or self._fpn_filter_scale > 1:
for i in range(self._min_level, self._max_level + 1):
depths.append(inputs[str(i)][-1] * 2)
else:
for _ in range(self._min_level, self._max_level + 1):
depths.append(minimum_depth)
minimum_depth *= 2
if self._fpn_input:
return depths
return list(reversed(depths))
def call(self, inputs):
outputs = dict()
layer_in = inputs[str(self._input)]
for level in self._iterator:
x_route, x = self.preprocessors[str(level)](layer_in)
outputs[str(level)] = x
if self._check(level):
x_next = inputs[str(self._key_shift(level))]
_, layer_in = self.resamples[str(
self._key_shift(level))]([x_route, x_next])
return outputs
@tf.keras.utils.register_keras_serializable(package='yolo')
class YoloDecoder(tf.keras.Model):
"""Darknet Backbone Decoder"""
def __init__(self,
input_specs,
use_fpn=False,
use_spatial_attention=False,
csp_stack=False,
fpn_depth=4,
fpn_filter_scale=1,
path_process_len=6,
max_level_process_len=None,
embed_spp=False,
activation='leaky',
use_sync_bn=False,
norm_momentum=0.99,
norm_epsilon=0.001,
kernel_initializer='glorot_uniform',
kernel_regularizer=None,
bias_regularizer=None,
**kwargs):
"""Yolo Decoder initialization function. A unified model that ties all decoder
components into a conditionally build YOLO decder.
Args:
input_specs: `dict[str, tf.InputSpec]`: input specs of each of the inputs
to the heads.
use_fpn: `bool`, use the FPN found in the YoloV4 model.
use_spatial_attention: `bool`, use the spatial attention module.
csp_stack: `bool`, CSPize the FPN.
fpn_depth: `int`, number of layers ot use in each FPN path
if you choose to use an FPN.
fpn_filter_scale: `int`, scaling factor for the FPN filters.
path_process_len: `int`, number of layers ot use in each Decoder path.
max_level_process_len: `int`, number of layers ot use in the largest
processing path, or the backbones largest output if it is different.
embed_spp: `bool`, use the SPP found in the YoloV3 and V4 model.
activation: `str`, the activation function to use typically leaky or mish.
use_sync_bn: if True, use synchronized batch normalization.
norm_momentum: `float`, normalization omentum for the moving average.
norm_epsilon: `float`, small float added to variance to avoid dividing by
zero.
kernel_initializer: kernel_initializer for convolutional layers.
kernel_regularizer: tf.keras.regularizers.Regularizer object for Conv2D.
bias_regularizer: tf.keras.regularizers.Regularizer object for Conv2d.
**kwargs: keyword arguments to be passed.
"""
self._input_specs = input_specs
self._use_fpn = use_fpn
self._fpn_depth = fpn_depth
self._path_process_len = path_process_len
self._max_level_process_len = max_level_process_len
self._embed_spp = embed_spp
self._activation = activation
self._use_sync_bn = use_sync_bn
self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon
self._kernel_initializer = kernel_initializer
self._kernel_regularizer = kernel_regularizer
self._bias_regularizer = bias_regularizer
self._base_config = dict(
use_spatial_attention=use_spatial_attention,
csp_stack=csp_stack,
activation=self._activation,
use_sync_bn=self._use_sync_bn,
fpn_filter_scale=fpn_filter_scale,
norm_momentum=self._norm_momentum,
norm_epsilon=self._norm_epsilon,
kernel_initializer=self._kernel_initializer,
kernel_regularizer=self._kernel_regularizer,
bias_regularizer=self._bias_regularizer)
self._decoder_config = dict(
path_process_len=self._path_process_len,
max_level_process_len=self._max_level_process_len,
embed_spp=self._embed_spp,
fpn_input=self._use_fpn,
**self._base_config)
inputs = {
key: tf.keras.layers.Input(shape=value[1:])
for key, value in input_specs.items()
}
if self._use_fpn:
inter_outs = YoloFPN(
fpn_depth=self._fpn_depth, **self._base_config)(
inputs)
outputs = YoloPAN(**self._decoder_config)(inter_outs)
else:
inter_outs = None
outputs = YoloPAN(**self._decoder_config)(inputs)
self._output_specs = {key: value.shape for key, value in outputs.items()}
super().__init__(inputs=inputs, outputs=outputs, name='YoloDecoder')
@property
def use_fpn(self):
return self._use_fpn
@property
def output_specs(self):
return self._output_specs
def get_config(self):
config = dict(
input_specs=self._input_specs,
use_fpn=self._use_fpn,
fpn_depth=self._fpn_depth,
**self._decoder_config)
return config
@classmethod
def from_config(cls, config, custom_objects=None):
return cls(**config)
official/vision/beta/projects/yolo/modeling/decoders/yolo_decoder_test.py 0 → 100644
View file @ 63719f08
# Copyright 2021 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# Lint as: python3
"""Tests for YOLO."""
# Import libraries
from absl.testing import parameterized
import numpy as np
import tensorflow as tf
from tensorflow.python.distribute import combinations
from tensorflow.python.distribute import strategy_combinations
from official.vision.beta.projects.yolo.modeling.decoders import yolo_decoder as decoders
class YoloDecoderTest(parameterized.TestCase, tf.test.TestCase):
@parameterized.parameters('1', '6spp', '6sppfpn', '6')
def test_network_creation(self, version):
"""Test creation of ResNet family models."""
tf.keras.backend.set_image_data_format('channels_last')
input_shape = {
'3': [1, 52, 52, 256],
'4': [1, 26, 26, 512],
'5': [1, 13, 13, 1024]
}
decoder = build_yolo_decoder(input_shape, version)
inputs = {}
for key in input_shape.keys():
inputs[key] = tf.ones(input_shape[key], dtype=tf.float32)
endpoints = decoder.call(inputs)
for key in endpoints.keys():
self.assertAllEqual(endpoints[key].shape.as_list(), input_shape[key])
@combinations.generate(
combinations.combine(
strategy=[
strategy_combinations.tpu_strategy,
strategy_combinations.one_device_strategy_gpu,
],
use_sync_bn=[False, True],
))
def test_sync_bn_multiple_devices(self, strategy, use_sync_bn):
"""Test for sync bn on TPU and GPU devices."""
tf.keras.backend.set_image_data_format('channels_last')
with strategy.scope():
input_shape = {
'3': [1, 52, 52, 256],
'4': [1, 26, 26, 512],
'5': [1, 13, 13, 1024]
}
decoder = build_yolo_decoder(input_shape, '6')
inputs = {}
for key in input_shape.keys():
inputs[key] = tf.ones(input_shape[key], dtype=tf.float32)
_ = decoder.call(inputs)
@parameterized.parameters(1, 3, 4)
def test_input_specs(self, input_dim):
"""Test different input feature dimensions."""
tf.keras.backend.set_image_data_format('channels_last')
input_shape = {
'3': [1, 52, 52, 256],
'4': [1, 26, 26, 512],
'5': [1, 13, 13, 1024]
}
decoder = build_yolo_decoder(input_shape, '6')
inputs = {}
for key in input_shape.keys():
inputs[key] = tf.ones(input_shape[key], dtype=tf.float32)
_ = decoder(inputs)
def test_serialize_deserialize(self):
"""Create a network object that sets all of its config options."""
tf.keras.backend.set_image_data_format('channels_last')
input_shape = {
'3': [1, 52, 52, 256],
'4': [1, 26, 26, 512],
'5': [1, 13, 13, 1024]
}
decoder = build_yolo_decoder(input_shape, '6')
inputs = {}
for key in input_shape.keys():
inputs[key] = tf.ones(input_shape[key], dtype=tf.float32)
_ = decoder(inputs)
config = decoder.get_config()
decoder_from_config = decoders.YoloDecoder.from_config(config)
self.assertAllEqual(decoder.get_config(), decoder_from_config.get_config())
def build_yolo_decoder(input_specs, type='1'):
if type == '1':
model = decoders.YoloDecoder(
input_specs=input_specs,
embed_spp=False,
use_fpn=False,
max_level_process_len=2,
path_process_len=1,
activation='mish')
elif type == '6spp':
model = decoders.YoloDecoder(
input_specs=input_specs,
embed_spp=True,
use_fpn=False,
max_level_process_len=None,
path_process_len=6,
activation='mish')
elif type == '6sppfpn':
model = decoders.YoloDecoder(
input_specs=input_specs,
embed_spp=True,
use_fpn=True,
max_level_process_len=None,
path_process_len=6,
activation='mish')
elif type == '6':
model = decoders.YoloDecoder(
input_specs=input_specs,
embed_spp=False,
use_fpn=False,
max_level_process_len=None,
path_process_len=6,
activation='mish')
else:
raise NotImplementedError(f"YOLO decoder test {type} not implemented.")
return model
if __name__ == '__main__':
tf.test.main()
official/vision/beta/projects/yolo/modeling/heads/yolo_head.py 0 → 100644
View file @ 63719f08
# Copyright 2021 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# Lint as: python3
import tensorflow as tf
from official.vision.beta.projects.yolo.modeling.layers import nn_blocks
class YoloHead(tf.keras.layers.Layer):
"""YOLO Prediction Head"""
def __init__(self,
min_level,
max_level,
classes=80,
boxes_per_level=3,
output_extras=0,
norm_momentum=0.99,
norm_epsilon=0.001,
kernel_initializer='glorot_uniform',
kernel_regularizer=None,
bias_regularizer=None,
activation=None,
**kwargs):
"""Yolo Prediction Head initialization function.
Args:
min_level: `int`, the minimum backbone output level.
max_level: `int`, the maximum backbone output level.
classes: `int`, number of classes per category.
boxes_per_level: `int`, number of boxes to predict per level.
output_extras: `int`, number of additional output channels that the head.
should predict for non-object detection and non-image classification
tasks.
norm_momentum: `float`, normalization momentum for the moving average.
norm_epsilon: `float`, small float added to variance to avoid dividing by
zero.
kernel_initializer: kernel_initializer for convolutional layers.
kernel_regularizer: tf.keras.regularizers.Regularizer object for Conv2D.
bias_regularizer: tf.keras.regularizers.Regularizer object for Conv2d.
activation: `str`, the activation function to use typically leaky or mish.
**kwargs: keyword arguments to be passed.
"""
super().__init__(**kwargs)
self._min_level = min_level
self._max_level = max_level
self._key_list = [
str(key) for key in range(self._min_level, self._max_level + 1)
]
self._classes = classes
self._boxes_per_level = boxes_per_level
self._output_extras = output_extras
self._output_conv = (classes + output_extras + 5) * boxes_per_level
self._base_config = dict(
activation=activation,
norm_momentum=norm_momentum,
norm_epsilon=norm_epsilon,
kernel_initializer=kernel_initializer,
kernel_regularizer=kernel_regularizer,
bias_regularizer=bias_regularizer)
self._conv_config = dict(
filters=self._output_conv,
kernel_size=(1, 1),
strides=(1, 1),
padding='same',
use_bn=False,
**self._base_config)
def build(self, input_shape):
self._head = dict()
for key in self._key_list:
self._head[key] = nn_blocks.ConvBN(**self._conv_config)
def call(self, inputs):
outputs = dict()
for key in self._key_list:
outputs[key] = self._head[key](inputs[key])
return outputs
@property
def output_depth(self):
return (self._classes + self._output_extras + 5) * self._boxes_per_level
@property
def num_boxes(self):
if self._min_level is None or self._max_level is None:
raise Exception(
'Model has to be built before number of boxes can be determined.')
return (self._max_level - self._min_level + 1) * self._boxes_per_level
def get_config(self):
config = dict(
min_level=self._min_level,
max_level=self._max_level,
classes=self._classes,
boxes_per_level=self._boxes_per_level,
output_extras=self._output_extras,
**self._base_config)
return config
@classmethod
def from_config(cls, config, custom_objects=None):
return cls(**config)
official/vision/beta/projects/yolo/modeling/heads/yolo_head_test.py 0 → 100644
View file @ 63719f08
# Copyright 2021 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# Lint as: python3
"""Tests for yolo heads."""
# Import libraries
from absl.testing import parameterized
import numpy as np
import tensorflow as tf
from tensorflow.python.distribute import combinations
from tensorflow.python.distribute import strategy_combinations
from official.vision.beta.projects.yolo.modeling.heads import yolo_head as heads
class YoloDecoderTest(parameterized.TestCase, tf.test.TestCase):
def test_network_creation(self):
"""Test creation of YOLO family models."""
tf.keras.backend.set_image_data_format('channels_last')
input_shape = {
'3': [1, 52, 52, 256],
'4': [1, 26, 26, 512],
'5': [1, 13, 13, 1024]
}
classes = 100
bps = 3
head = heads.YoloHead(3, 5, classes=classes, boxes_per_level=bps)
inputs = {}
for key in input_shape.keys():
inputs[key] = tf.ones(input_shape[key], dtype=tf.float32)
endpoints = head(inputs)
# print(endpoints)
for key in endpoints.keys():
expected_input_shape = input_shape[key]
expected_input_shape[-1] = (classes + 5) * bps
self.assertAllEqual(endpoints[key].shape.as_list(), expected_input_shape)
def test_serialize_deserialize(self):
# Create a network object that sets all of its config options.
tf.keras.backend.set_image_data_format('channels_last')
input_shape = {
'3': [1, 52, 52, 256],
'4': [1, 26, 26, 512],
'5': [1, 13, 13, 1024]
}
classes = 100
bps = 3
head = heads.YoloHead(3, 5, classes=classes, boxes_per_level=bps)
inputs = {}
for key in input_shape.keys():
inputs[key] = tf.ones(input_shape[key], dtype=tf.float32)
_ = head(inputs)
configs = head.get_config()
head_from_config = heads.YoloHead.from_config(configs)
self.assertAllEqual(head.get_config(), head_from_config.get_config())
if __name__ == '__main__':
tf.test.main()
official/vision/beta/projects/yolo/modeling/layers/nn_blocks.py
View file @ 63719f08
...@@ -15,62 +15,64 @@ ...@@ -15,62 +15,64 @@
# Lint as: python3 # Lint as: python3
"""Contains common building blocks for yolo neural networks.""" """Contains common building blocks for yolo neural networks."""
from typing import Callable, List from typing import Callable, List
import tensorflow as tf import tensorflow as tf
from official.modeling import tf_utils from official.modeling import tf_utils
from official.vision.beta.ops import spatial_transform_ops
@tf.keras.utils.register_keras_serializable(package="yolo") @tf.keras.utils.register_keras_serializable(package='yolo')
class Identity(tf.keras.layers.Layer): class Identity(tf.keras.layers.Layer):
def __init__(self, **kwargs):
super().__init__(**kwargs)
def call(self, inputs): def call(self, inputs):
return inputs return inputs
@tf.keras.utils.register_keras_serializable(package="yolo") @tf.keras.utils.register_keras_serializable(package='yolo')
class ConvBN(tf.keras.layers.Layer): class ConvBN(tf.keras.layers.Layer):
"""Modified Convolution layer to match that of the DarkNet Library. """
Modified Convolution layer to match that of the Darknet Library.
The Layer is a standards combination of Conv BatchNorm Activation, The Layer is a standards combination of Conv BatchNorm Activation,
however, the use of bias in the conv is determined by the use of batch norm. however, the use of bias in the conv is determined by the use of batch
normalization.
Cross Stage Partial networks (CSPNets) were proposed in: Cross Stage Partial networks (CSPNets) were proposed in:
[1] Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu, Ping-Yang [1] Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu,
Chen, Jun-Wei Hsieh. Ping-Yang Chen, Jun-Wei Hsieh
CSPNet: A New Backbone that can Enhance Learning Capability of CNN. CSPNet: A New Backbone that can Enhance Learning Capability of CNN.
arXiv:1911.11929 arXiv:1911.11929
""" """
def __init__(self, def __init__(self,
filters=1, filters=1,
kernel_size=(1, 1), kernel_size=(1, 1),
strides=(1, 1), strides=(1, 1),
padding="same", padding='same',
dilation_rate=(1, 1), dilation_rate=(1, 1),
kernel_initializer="glorot_uniform", kernel_initializer='glorot_uniform',
bias_initializer="zeros", bias_initializer='zeros',
kernel_regularizer=None,
bias_regularizer=None, bias_regularizer=None,
kernel_regularizer=None,
use_bn=True, use_bn=True,
use_sync_bn=False, use_sync_bn=False,
norm_momentum=0.99, norm_momentum=0.99,
norm_epsilon=0.001, norm_epsilon=0.001,
activation="leaky", activation='leaky',
leaky_alpha=0.1, leaky_alpha=0.1,
**kwargs): **kwargs):
"""Initializes ConvBN layer. """
Args: Args:
filters: integer for output depth, or the number of features to learn filters: integer for output depth, or the number of features to learn.
kernel_size: integer or tuple for the shape of the weight matrix or kernel kernel_size: integer or tuple for the shape of the weight matrix or kernel
to learn. to learn.
strides: integer of tuple how much to move the kernel after each kernel strides: integer of tuple how much to move the kernel after each kernel
use padding: string 'valid' or 'same', if same, then pad the image, else use.
do not. padding: string 'valid' or 'same', if same, then pad the image, else do
padding: `str`, padding method for conv layers. not.
dilation_rate: tuple to indicate how much to modulate kernel weights and dialtion_rate: tuple to indicate how much to modulate kernel weights and
how many pixels in a feature map to skip. how many pixels in a feature map to skip.
kernel_initializer: string to indicate which function to use to initialize kernel_initializer: string to indicate which function to use to initialize
weights. weights.
bias_initializer: string to indicate which function to use to initialize bias_initializer: string to indicate which function to use to initialize
...@@ -80,14 +82,17 @@ class ConvBN(tf.keras.layers.Layer): ...@@ -80,14 +82,17 @@ class ConvBN(tf.keras.layers.Layer):
bias_regularizer: string to indicate which function to use to regularizer bias_regularizer: string to indicate which function to use to regularizer
bias. bias.
use_bn: boolean for whether to use batch normalization. use_bn: boolean for whether to use batch normalization.
use_sync_bn: boolean for whether sync batch normalization. use_sync_bn: boolean for whether sync batch normalization statistics
norm_momentum: float for moment to use for batch normalization of all batch norm layers to the models global statistics
norm_epsilon: float for batch normalization epsilon (across all input batches).
norm_momentum: float for moment to use for batch normalization.
norm_epsilon: float for batch normalization epsilon.
activation: string or None for activation function to use in layer, activation: string or None for activation function to use in layer,
if None activation is replaced by linear. if None activation is replaced by linear.
leaky_alpha: float to use as alpha if activation function is leaky. leaky_alpha: float to use as alpha if activation function is leaky.
**kwargs: Keyword Arguments **kwargs: Keyword Arguments.
""" """
# convolution params # convolution params
self._filters = filters self._filters = filters
self._kernel_size = kernel_size self._kernel_size = kernel_size
...@@ -97,15 +102,16 @@ class ConvBN(tf.keras.layers.Layer): ...@@ -97,15 +102,16 @@ class ConvBN(tf.keras.layers.Layer):
self._kernel_initializer = kernel_initializer self._kernel_initializer = kernel_initializer
self._bias_initializer = bias_initializer self._bias_initializer = bias_initializer
self._kernel_regularizer = kernel_regularizer self._kernel_regularizer = kernel_regularizer
self._bias_regularizer = bias_regularizer self._bias_regularizer = bias_regularizer
# batch normalization params # batch normalization params
self._use_bn = use_bn self._use_bn = use_bn
self._use_sync_bn = use_sync_bn self._use_sync_bn = use_sync_bn
self._norm_moment = norm_momentum self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon self._norm_epsilon = norm_epsilon
if tf.keras.backend.image_data_format() == "channels_last": if tf.keras.backend.image_data_format() == 'channels_last':
# format: (batch_size, height, width, channels) # format: (batch_size, height, width, channels)
self._bn_axis = -1 self._bn_axis = -1
else: else:
...@@ -116,7 +122,7 @@ class ConvBN(tf.keras.layers.Layer): ...@@ -116,7 +122,7 @@ class ConvBN(tf.keras.layers.Layer):
self._activation = activation self._activation = activation
self._leaky_alpha = leaky_alpha self._leaky_alpha = leaky_alpha
super(ConvBN, self).__init__(**kwargs) super().__init__(**kwargs)
def build(self, input_shape): def build(self, input_shape):
use_bias = not self._use_bn use_bias = not self._use_bn
...@@ -136,101 +142,101 @@ class ConvBN(tf.keras.layers.Layer): ...@@ -136,101 +142,101 @@ class ConvBN(tf.keras.layers.Layer):
if self._use_bn: if self._use_bn:
if self._use_sync_bn: if self._use_sync_bn:
self.bn = tf.keras.layers.experimental.SyncBatchNormalization( self.bn = tf.keras.layers.experimental.SyncBatchNormalization(
momentum=self._norm_moment, momentum=self._norm_momentum,
epsilon=self._norm_epsilon, epsilon=self._norm_epsilon,
axis=self._bn_axis) axis=self._bn_axis)
else: else:
self.bn = tf.keras.layers.BatchNormalization( self.bn = tf.keras.layers.BatchNormalization(
momentum=self._norm_moment, momentum=self._norm_momentum,
epsilon=self._norm_epsilon, epsilon=self._norm_epsilon,
axis=self._bn_axis) axis=self._bn_axis)
else:
self.bn = Identity()
if self._activation == "leaky": if self._activation == 'leaky':
self._activation_fn = tf.keras.layers.LeakyReLU(alpha=self._leaky_alpha) self._activation_fn = tf.keras.layers.LeakyReLU(alpha=self._leaky_alpha)
elif self._activation == "mish": elif self._activation == 'mish':
self._activation_fn = lambda x: x * tf.math.tanh(tf.math.softplus(x)) self._activation_fn = lambda x: x * tf.math.tanh(tf.math.softplus(x))
else: else:
self._activation_fn = tf_utils.get_activation(self._activation) self._activation_fn = tf_utils.get_activation(self._activation)
def call(self, x): def call(self, x):
x = self.conv(x) x = self.conv(x)
x = self.bn(x) if self._use_bn:
x = self.bn(x)
x = self._activation_fn(x) x = self._activation_fn(x)
return x return x
def get_config(self): def get_config(self):
# used to store/share parameters to reconstruct the model # used to store/share parameters to reconstruct the model
layer_config = { layer_config = {
"filters": self._filters, 'filters': self._filters,
"kernel_size": self._kernel_size, 'kernel_size': self._kernel_size,
"strides": self._strides, 'strides': self._strides,
"padding": self._padding, 'padding': self._padding,
"dilation_rate": self._dilation_rate, 'dilation_rate': self._dilation_rate,
"kernel_initializer": self._kernel_initializer, 'kernel_initializer': self._kernel_initializer,
"bias_initializer": self._bias_initializer, 'bias_initializer': self._bias_initializer,
"bias_regularizer": self._bias_regularizer, 'bias_regularizer': self._bias_regularizer,
"kernel_regularizer": self._kernel_regularizer, 'kernel_regularizer': self._kernel_regularizer,
"use_bn": self._use_bn, 'use_bn': self._use_bn,
"use_sync_bn": self._use_sync_bn, 'use_sync_bn': self._use_sync_bn,
"norm_moment": self._norm_moment, 'norm_momentum': self._norm_momentum,
"norm_epsilon": self._norm_epsilon, 'norm_epsilon': self._norm_epsilon,
"activation": self._activation, 'activation': self._activation,
"leaky_alpha": self._leaky_alpha 'leaky_alpha': self._leaky_alpha
} }
layer_config.update(super(ConvBN, self).get_config()) layer_config.update(super().get_config())
return layer_config return layer_config
def __repr__(self):
return repr(self.get_config())
@tf.keras.utils.register_keras_serializable(package="yolo") @tf.keras.utils.register_keras_serializable(package='yolo')
class DarkResidual(tf.keras.layers.Layer): class DarkResidual(tf.keras.layers.Layer):
"""DarkNet block with Residual connection for Yolo v3 Backbone. """
Darknet block with Residual connection for Yolo v3 Backbone
""" """
def __init__(self, def __init__(self,
filters=1, filters=1,
filter_scale=2, filter_scale=2,
kernel_initializer="glorot_uniform", dilation_rate=1,
bias_initializer="zeros", kernel_initializer='glorot_uniform',
bias_initializer='zeros',
kernel_regularizer=None, kernel_regularizer=None,
bias_regularizer=None, bias_regularizer=None,
use_bn=True, use_bn=True,
use_sync_bn=False, use_sync_bn=False,
norm_momentum=0.99, norm_momentum=0.99,
norm_epsilon=0.001, norm_epsilon=0.001,
activation="leaky", activation='leaky',
leaky_alpha=0.1, leaky_alpha=0.1,
sc_activation="linear", sc_activation='linear',
downsample=False, downsample=False,
**kwargs): **kwargs):
"""Initializes DarkResidual. """
Args: Args:
filters: integer for output depth, or the number of features to learn. filters: integer for output depth, or the number of features to learn.
filter_scale: `int`, scale factor for number of filters.
kernel_initializer: string to indicate which function to use to initialize kernel_initializer: string to indicate which function to use to initialize
weights weights.
bias_initializer: string to indicate which function to use to initialize bias_initializer: string to indicate which function to use to initialize
bias bias.
kernel_regularizer: string to indicate which function to use to kernel_regularizer: string to indicate which function to use to
regularizer weights regularizer weights.
bias_regularizer: string to indicate which function to use to regularizer bias_regularizer: string to indicate which function to use to regularizer
bias bias.
use_bn: boolean for whether to use batch normalization use_bn: boolean for whether to use batch normalization.
use_sync_bn: boolean for whether sync batch normalization. use_sync_bn: boolean for whether sync batch normalization statistics.
norm_momentum: float for moment to use for batch normalization of all batch norm layers to the models global statistics
norm_epsilon: float for batch normalization epsilon (across all input batches).
activation: string for activation function to use in conv layers. norm_momentum: float for moment to use for batch normalization.
leaky_alpha: float to use as alpha if activation function is leaky norm_epsilon: float for batch normalization epsilon.
sc_activation: string for activation function to use in layer conv_activation: string or None for activation function to use in layer,
if None activation is replaced by linear.
leaky_alpha: float to use as alpha if activation function is leaky.
sc_activation: string for activation function to use in layer.
downsample: boolean for if image input is larger than layer output, set downsample: boolean for if image input is larger than layer output, set
downsample to True so the dimensions are forced to match downsample to True so the dimensions are forced to match.
**kwargs: Keyword Arguments **kwargs: Keyword Arguments.
""" """
# downsample # downsample
self._downsample = downsample self._downsample = downsample
...@@ -245,8 +251,10 @@ class DarkResidual(tf.keras.layers.Layer): ...@@ -245,8 +251,10 @@ class DarkResidual(tf.keras.layers.Layer):
self._kernel_regularizer = kernel_regularizer self._kernel_regularizer = kernel_regularizer
# normal params # normal params
self._norm_moment = norm_momentum self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon self._norm_epsilon = norm_epsilon
self._dilation_rate = dilation_rate if isinstance(dilation_rate,
int) else dilation_rate[0]
# activation params # activation params
self._conv_activation = activation self._conv_activation = activation
...@@ -256,138 +264,150 @@ class DarkResidual(tf.keras.layers.Layer): ...@@ -256,138 +264,150 @@ class DarkResidual(tf.keras.layers.Layer):
super().__init__(**kwargs) super().__init__(**kwargs)
def build(self, input_shape): def build(self, input_shape):
self._dark_conv_args = { dark_conv_args = {
"kernel_initializer": self._kernel_initializer, 'kernel_initializer': self._kernel_initializer,
"bias_initializer": self._bias_initializer, 'bias_initializer': self._bias_initializer,
"bias_regularizer": self._bias_regularizer, 'bias_regularizer': self._bias_regularizer,
"use_bn": self._use_bn, 'use_bn': self._use_bn,
"use_sync_bn": self._use_sync_bn, 'use_sync_bn': self._use_sync_bn,
"norm_momentum": self._norm_moment, 'norm_momentum': self._norm_momentum,
"norm_epsilon": self._norm_epsilon, 'norm_epsilon': self._norm_epsilon,
"activation": self._conv_activation, 'activation': self._conv_activation,
"kernel_regularizer": self._kernel_regularizer, 'kernel_regularizer': self._kernel_regularizer,
"leaky_alpha": self._leaky_alpha 'leaky_alpha': self._leaky_alpha
} }
if self._downsample: if self._downsample:
if self._dilation_rate > 1:
dilation_rate = 1
if self._dilation_rate // 2 > 0:
dilation_rate = self._dilation_rate // 2
down_stride = 1
else:
dilation_rate = 1
down_stride = 2
self._dconv = ConvBN( self._dconv = ConvBN(
filters=self._filters, filters=self._filters,
kernel_size=(3, 3), kernel_size=(3, 3),
strides=(2, 2), strides=down_stride,
padding="same", dilation_rate=dilation_rate,
**self._dark_conv_args) padding='same',
else: **dark_conv_args)
self._dconv = Identity()
self._conv1 = ConvBN( self._conv1 = ConvBN(
filters=self._filters // self._filter_scale, filters=self._filters // self._filter_scale,
kernel_size=(1, 1), kernel_size=(1, 1),
strides=(1, 1), strides=(1, 1),
padding="same", padding='same',
**self._dark_conv_args) **dark_conv_args)
self._conv2 = ConvBN( self._conv2 = ConvBN(
filters=self._filters, filters=self._filters,
kernel_size=(3, 3), kernel_size=(3, 3),
strides=(1, 1), strides=(1, 1),
padding="same", dilation_rate=self._dilation_rate,
**self._dark_conv_args) padding='same',
**dark_conv_args)
self._shortcut = tf.keras.layers.Add() self._shortcut = tf.keras.layers.Add()
if self._sc_activation == "leaky": if self._sc_activation == 'leaky':
self._activation_fn = tf.keras.layers.LeakyReLU( self._activation_fn = tf.keras.layers.LeakyReLU(alpha=self._leaky_alpha)
alpha=self._leaky_alpha) elif self._sc_activation == 'mish':
elif self._sc_activation == "mish":
self._activation_fn = lambda x: x * tf.math.tanh(tf.math.softplus(x)) self._activation_fn = lambda x: x * tf.math.tanh(tf.math.softplus(x))
else: else:
self._activation_fn = tf_utils.get_activation(self._sc_activation) self._activation_fn = tf_utils.get_activation(
self._sc_activation
)
super().build(input_shape) super().build(input_shape)
def call(self, inputs): def call(self, inputs, training=None):
shortcut = self._dconv(inputs) if self._downsample:
x = self._conv1(shortcut) inputs = self._dconv(inputs)
x = self._conv1(inputs)
x = self._conv2(x) x = self._conv2(x)
x = self._shortcut([x, shortcut]) x = self._shortcut([x, inputs])
return self._activation_fn(x) return self._activation_fn(x)
def get_config(self): def get_config(self):
# used to store/share parameters to reconstruct the model # used to store/share parameters to reconstruct the model
layer_config = { layer_config = {
"filters": self._filters, 'filters': self._filters,
"kernel_initializer": self._kernel_initializer, 'kernel_initializer': self._kernel_initializer,
"bias_initializer": self._bias_initializer, 'bias_initializer': self._bias_initializer,
"kernel_regularizer": self._kernel_regularizer, 'kernel_regularizer': self._kernel_regularizer,
"use_bn": self._use_bn, 'dilation_rate': self._dilation_rate,
"use_sync_bn": self._use_sync_bn, 'use_bn': self._use_bn,
"norm_moment": self._norm_moment, 'use_sync_bn': self._use_sync_bn,
"norm_epsilon": self._norm_epsilon, 'norm_momentum': self._norm_momentum,
"activation": self._conv_activation, 'norm_epsilon': self._norm_epsilon,
"leaky_alpha": self._leaky_alpha, 'activation': self._conv_activation,
"sc_activation": self._sc_activation, 'leaky_alpha': self._leaky_alpha,
"downsample": self._downsample 'sc_activation': self._sc_activation,
'downsample': self._downsample,
} }
layer_config.update(super().get_config()) layer_config.update(super().get_config())
return layer_config return layer_config
@tf.keras.utils.register_keras_serializable(package="yolo") @tf.keras.utils.register_keras_serializable(package='yolo')
class CSPTiny(tf.keras.layers.Layer): class CSPTiny(tf.keras.layers.Layer):
"""A Small size convolution block proposed in the CSPNet. """
A Small size convolution block proposed in the CSPNet. The layer uses
The layer uses shortcuts, routing(concatnation), and feature grouping shortcuts, routing(concatnation), and feature grouping in order to improve
in order to improve gradient variablity and allow for high efficency, low gradient variablity and allow for high efficency, low power residual learning
power residual learning for small networtf.keras. for small networtf.keras.
Cross Stage Partial networks (CSPNets) were proposed in: Cross Stage Partial networks (CSPNets) were proposed in:
[1] Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu, Ping-Yang [1] Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu,
Chen, Jun-Wei Hsieh Ping-Yang Chen, Jun-Wei Hsieh
CSPNet: A New Backbone that can Enhance Learning Capability of CNN. CSPNet: A New Backbone that can Enhance Learning Capability of CNN.
arXiv:1911.11929 arXiv:1911.11929
""" """
def __init__(self, def __init__(self,
filters=1, filters=1,
kernel_initializer="glorot_uniform", kernel_initializer='glorot_uniform',
bias_initializer="zeros", bias_initializer='zeros',
kernel_regularizer=None,
bias_regularizer=None, bias_regularizer=None,
kernel_regularizer=None,
use_bn=True, use_bn=True,
dilation_rate=1,
use_sync_bn=False, use_sync_bn=False,
group_id=1, group_id=1,
groups=2, groups=2,
norm_momentum=0.99, norm_momentum=0.99,
norm_epsilon=0.001, norm_epsilon=0.001,
activation="leaky", activation='leaky',
downsample=True, downsample=True,
leaky_alpha=0.1, leaky_alpha=0.1,
**kwargs): **kwargs):
"""Initializes CSPTiny. """
Args: Args:
filters: integer for output depth, or the number of features to learn filters: integer for output depth, or the number of features to learn.
kernel_initializer: string to indicate which function to use to initialize kernel_initializer: string to indicate which function to use to initialize
weights weights.
bias_initializer: string to indicate which function to use to initialize bias_initializer: string to indicate which function to use to initialize
bias bias.
use_bn: boolean for whether to use batch normalization.
kernel_regularizer: string to indicate which function to use to kernel_regularizer: string to indicate which function to use to
regularizer weights regularizer weights.
bias_regularizer: string to indicate which function to use to regularizer bias_regularizer: string to indicate which function to use to regularizer
bias bias.
use_bn: boolean for whether to use batch normalization use_sync_bn: boolean for whether sync batch normalization statistics
use_sync_bn: boolean for whether sync batch normalization statistics of of all batch norm layers to the models global statistics
all batch norm layers to the models global statistics (across all input (across all input batches).
batches) group_id: integer for which group of features to pass through the csp
group_id: integer for which group of features to pass through the csp tiny tiny stack.
stack.
groups: integer for how many splits there should be in the convolution groups: integer for how many splits there should be in the convolution
feature stack output feature stack output.
norm_momentum: float for moment to use for batch normalization norm_momentum: float for moment to use for batch normalization.
norm_epsilon: float for batch normalization epsilon norm_epsilon: float for batch normalization epsilon.
activation: string or None for activation function to use in layer, conv_activation: string or None for activation function to use in layer,
if None activation is replaced by linear if None activation is replaced by linear.
leaky_alpha: float to use as alpha if activation function is leaky.
sc_activation: string for activation function to use in layer.
downsample: boolean for if image input is larger than layer output, set downsample: boolean for if image input is larger than layer output, set
downsample to True so the dimensions are forced to match downsample to True so the dimensions are forced to match.
leaky_alpha: float to use as alpha if activation function is leaky **kwargs: Keyword Arguments.
**kwargs: Keyword Arguments
""" """
# ConvBN params # ConvBN params
...@@ -396,6 +416,7 @@ class CSPTiny(tf.keras.layers.Layer): ...@@ -396,6 +416,7 @@ class CSPTiny(tf.keras.layers.Layer):
self._bias_initializer = bias_initializer self._bias_initializer = bias_initializer
self._bias_regularizer = bias_regularizer self._bias_regularizer = bias_regularizer
self._use_bn = use_bn self._use_bn = use_bn
self._dilation_rate = dilation_rate
self._use_sync_bn = use_sync_bn self._use_sync_bn = use_sync_bn
self._kernel_regularizer = kernel_regularizer self._kernel_regularizer = kernel_regularizer
self._groups = groups self._groups = groups
...@@ -403,7 +424,7 @@ class CSPTiny(tf.keras.layers.Layer): ...@@ -403,7 +424,7 @@ class CSPTiny(tf.keras.layers.Layer):
self._downsample = downsample self._downsample = downsample
# normal params # normal params
self._norm_moment = norm_momentum self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon self._norm_epsilon = norm_epsilon
# activation params # activation params
...@@ -413,37 +434,37 @@ class CSPTiny(tf.keras.layers.Layer): ...@@ -413,37 +434,37 @@ class CSPTiny(tf.keras.layers.Layer):
super().__init__(**kwargs) super().__init__(**kwargs)
def build(self, input_shape): def build(self, input_shape):
self._dark_conv_args = { dark_conv_args = {
"kernel_initializer": self._kernel_initializer, 'kernel_initializer': self._kernel_initializer,
"bias_initializer": self._bias_initializer, 'bias_initializer': self._bias_initializer,
"bias_regularizer": self._bias_regularizer, 'bias_regularizer': self._bias_regularizer,
"use_bn": self._use_bn, 'use_bn': self._use_bn,
"use_sync_bn": self._use_sync_bn, 'use_sync_bn': self._use_sync_bn,
"norm_momentum": self._norm_moment, 'norm_momentum': self._norm_momentum,
"norm_epsilon": self._norm_epsilon, 'norm_epsilon': self._norm_epsilon,
"activation": self._conv_activation, 'activation': self._conv_activation,
"kernel_regularizer": self._kernel_regularizer, 'kernel_regularizer': self._kernel_regularizer,
"leaky_alpha": self._leaky_alpha 'leaky_alpha': self._leaky_alpha
} }
self._convlayer1 = ConvBN( self._convlayer1 = ConvBN(
filters=self._filters, filters=self._filters,
kernel_size=(3, 3), kernel_size=(3, 3),
strides=(1, 1), strides=(1, 1),
padding="same", padding='same',
**self._dark_conv_args) **dark_conv_args)
self._convlayer2 = ConvBN( self._convlayer2 = ConvBN(
filters=self._filters // 2, filters=self._filters // 2,
kernel_size=(3, 3), kernel_size=(3, 3),
strides=(1, 1), strides=(1, 1),
padding="same", padding='same',
kernel_initializer=self._kernel_initializer, kernel_initializer=self._kernel_initializer,
bias_initializer=self._bias_initializer, bias_initializer=self._bias_initializer,
bias_regularizer=self._bias_regularizer, bias_regularizer=self._bias_regularizer,
kernel_regularizer=self._kernel_regularizer, kernel_regularizer=self._kernel_regularizer,
use_bn=self._use_bn, use_bn=self._use_bn,
use_sync_bn=self._use_sync_bn, use_sync_bn=self._use_sync_bn,
norm_momentum=self._norm_moment, norm_momentum=self._norm_momentum,
norm_epsilon=self._norm_epsilon, norm_epsilon=self._norm_epsilon,
activation=self._conv_activation, activation=self._conv_activation,
leaky_alpha=self._leaky_alpha) leaky_alpha=self._leaky_alpha)
...@@ -452,22 +473,23 @@ class CSPTiny(tf.keras.layers.Layer): ...@@ -452,22 +473,23 @@ class CSPTiny(tf.keras.layers.Layer):
filters=self._filters // 2, filters=self._filters // 2,
kernel_size=(3, 3), kernel_size=(3, 3),
strides=(1, 1), strides=(1, 1),
padding="same", padding='same',
**self._dark_conv_args) **dark_conv_args)
self._convlayer4 = ConvBN( self._convlayer4 = ConvBN(
filters=self._filters, filters=self._filters,
kernel_size=(1, 1), kernel_size=(1, 1),
strides=(1, 1), strides=(1, 1),
padding="same", padding='same',
**self._dark_conv_args) **dark_conv_args)
self._maxpool = tf.keras.layers.MaxPool2D( if self._downsample:
pool_size=2, strides=2, padding="same", data_format=None) self._maxpool = tf.keras.layers.MaxPool2D(
pool_size=2, strides=2, padding='same', data_format=None)
super().build(input_shape) super().build(input_shape)
def call(self, inputs): def call(self, inputs, training=None):
x1 = self._convlayer1(inputs) x1 = self._convlayer1(inputs)
x1_group = tf.split(x1, self._groups, axis=-1)[self._group_id] x1_group = tf.split(x1, self._groups, axis=-1)[self._group_id]
x2 = self._convlayer2(x1_group) # grouping x2 = self._convlayer2(x1_group) # grouping
...@@ -479,74 +501,55 @@ class CSPTiny(tf.keras.layers.Layer): ...@@ -479,74 +501,55 @@ class CSPTiny(tf.keras.layers.Layer):
x = self._maxpool(x) x = self._maxpool(x)
return x, x5 return x, x5
def get_config(self):
# used to store/share parameters to reconsturct the model
layer_config = {
"filters": self._filters,
"strides": self._strides,
"kernel_initializer": self._kernel_initializer,
"bias_initializer": self._bias_initializer,
"kernel_regularizer": self._kernel_regularizer,
"use_bn": self._use_bn,
"use_sync_bn": self._use_sync_bn,
"norm_moment": self._norm_moment,
"norm_epsilon": self._norm_epsilon,
"activation": self._conv_activation,
"leaky_alpha": self._leaky_alpha,
"sc_activation": self._sc_activation,
}
layer_config.update(super().get_config())
return layer_config
@tf.keras.utils.register_keras_serializable(package="yolo") @tf.keras.utils.register_keras_serializable(package='yolo')
class CSPRoute(tf.keras.layers.Layer): class CSPRoute(tf.keras.layers.Layer):
"""Down sampling layer to take the place of down sampleing. """
Down sampling layer to take the place of down sampleing done in Residual
It is applied in Residual networks. This is the first of 2 layers needed to networks. This is the first of 2 layers needed to convert any Residual Network
convert any Residual Network model to a CSPNet. At the start of a new level model to a CSPNet. At the start of a new level change, this CSPRoute layer
change, this CSPRoute layer creates a learned identity that will act as a creates a learned identity that will act as a cross stage connection,
cross stage connection, that is used to inform the inputs to the next stage. that is used to inform the inputs to the next stage. It is called cross stage
It is called cross stage partial because the number of filters required in partial because the number of filters required in every intermitent Residual
every intermitent Residual layer is reduced by half. The sister layer will layer is reduced by half. The sister layer will take the partial generated by
take the partial generated by this layer and concatnate it with the output of this layer and concatnate it with the output of the final residual layer in
the final residual layer in the stack to create a fully feature level output. the stack to create a fully feature level output. This concatnation merges the
This concatnation merges the partial blocks of 2 levels as input to the next partial blocks of 2 levels as input to the next allowing the gradients of each
allowing the gradients of each level to be more unique, and reducing the level to be more unique, and reducing the number of parameters required by
number of parameters required by each level by 50% while keeping accuracy each level by 50% while keeping accuracy consistent.
consistent.
Cross Stage Partial networks (CSPNets) were proposed in: Cross Stage Partial networks (CSPNets) were proposed in:
[1] Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu, Ping-Yang [1] Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu,
Chen, Jun-Wei Hsieh. Ping-Yang Chen, Jun-Wei Hsieh
CSPNet: A New Backbone that can Enhance Learning Capability of CNN. CSPNet: A New Backbone that can Enhance Learning Capability of CNN.
arXiv:1911.11929 arXiv:1911.11929
""" """
def __init__(self, def __init__(self,
filters, filters,
filter_scale=2, filter_scale=2,
activation="mish", activation='mish',
downsample=True, kernel_initializer='glorot_uniform',
kernel_initializer="glorot_uniform", bias_initializer='zeros',
bias_initializer="zeros",
kernel_regularizer=None,
bias_regularizer=None, bias_regularizer=None,
kernel_regularizer=None,
dilation_rate=1,
use_bn=True, use_bn=True,
use_sync_bn=False, use_sync_bn=False,
norm_momentum=0.99, norm_momentum=0.99,
norm_epsilon=0.001, norm_epsilon=0.001,
downsample=True,
leaky_alpha=0.1,
**kwargs): **kwargs):
"""Initializes CSPRoute. """
Args: Args:
filters: integer for output depth, or the number of features to learn filters: integer for output depth, or the number of features to learn
filter_scale: integer dicating (filters//2) or the number of filters in filter_scale: integer dicating (filters//2) or the number of filters in
the partial feature stack. the partial feature stack.
activation: string for activation function to use in layer
downsample: down_sample the input. downsample: down_sample the input.
kernel_initializer: string to indicate which function to use to initialize activation: string for activation function to use in layer.
weights. kernel_initializer: string to indicate which function to use to
initialize weights.
bias_initializer: string to indicate which function to use to initialize bias_initializer: string to indicate which function to use to initialize
bias. bias.
kernel_regularizer: string to indicate which function to use to kernel_regularizer: string to indicate which function to use to
...@@ -554,99 +557,117 @@ class CSPRoute(tf.keras.layers.Layer): ...@@ -554,99 +557,117 @@ class CSPRoute(tf.keras.layers.Layer):
bias_regularizer: string to indicate which function to use to regularizer bias_regularizer: string to indicate which function to use to regularizer
bias. bias.
use_bn: boolean for whether to use batch normalization. use_bn: boolean for whether to use batch normalization.
use_sync_bn: boolean for whether sync batch normalization. use_sync_bn: boolean for whether sync batch normalization statistics
norm_momentum: float for moment to use for batch normalization of all batch norm layers to the models global statistics
norm_epsilon: float for batch normalization epsilon (across all input batches).
**kwargs: Keyword Arguments norm_momentum: float for moment to use for batch normalization.
norm_epsilon: float for batch normalization epsilon.
**kwargs: Keyword Arguments.
""" """
super().__init__(**kwargs) super().__init__(**kwargs)
# Layer params. # layer params
self._filters = filters self._filters = filters
self._filter_scale = filter_scale self._filter_scale = filter_scale
self._activation = activation self._activation = activation
# Convoultion params. # convoultion params
self._kernel_initializer = kernel_initializer self._kernel_initializer = kernel_initializer
self._bias_initializer = bias_initializer self._bias_initializer = bias_initializer
self._kernel_regularizer = kernel_regularizer self._kernel_regularizer = kernel_regularizer
self._bias_regularizer = bias_regularizer self._bias_regularizer = bias_regularizer
self._dilation_rate = dilation_rate
self._use_bn = use_bn self._use_bn = use_bn
self._use_sync_bn = use_sync_bn self._use_sync_bn = use_sync_bn
self._norm_moment = norm_momentum self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon self._norm_epsilon = norm_epsilon
self._downsample = downsample self._downsample = downsample
self._leaky_alpha = leaky_alpha
def build(self, input_shape): def build(self, input_shape):
self._dark_conv_args = { dark_conv_args = {
"kernel_initializer": self._kernel_initializer, 'kernel_initializer': self._kernel_initializer,
"bias_initializer": self._bias_initializer, 'bias_initializer': self._bias_initializer,
"bias_regularizer": self._bias_regularizer, 'bias_regularizer': self._bias_regularizer,
"use_bn": self._use_bn, 'use_bn': self._use_bn,
"use_sync_bn": self._use_sync_bn, 'use_sync_bn': self._use_sync_bn,
"norm_momentum": self._norm_moment, 'norm_momentum': self._norm_momentum,
"norm_epsilon": self._norm_epsilon, 'norm_epsilon': self._norm_epsilon,
"activation": self._activation, 'activation': self._activation,
"kernel_regularizer": self._kernel_regularizer, 'kernel_regularizer': self._kernel_regularizer,
'leaky_alpha': self._leaky_alpha,
} }
if self._downsample: if self._downsample:
self._conv1 = ConvBN(filters=self._filters, if self._dilation_rate > 1:
kernel_size=(3, 3), dilation_rate = 1
strides=(2, 2), if self._dilation_rate // 2 > 0:
**self._dark_conv_args) dilation_rate = self._dilation_rate // 2
else: down_stride = 1
self._conv1 = ConvBN(filters=self._filters, else:
kernel_size=(3, 3), dilation_rate = 1
strides=(1, 1), down_stride = 2
**self._dark_conv_args)
self._conv2 = ConvBN(filters=self._filters // self._filter_scale,
kernel_size=(1, 1),
strides=(1, 1),
**self._dark_conv_args)
self._conv3 = ConvBN(filters=self._filters // self._filter_scale,
kernel_size=(1, 1),
strides=(1, 1),
**self._dark_conv_args)
def call(self, inputs): self._conv1 = ConvBN(
x = self._conv1(inputs) filters=self._filters,
y = self._conv2(x) kernel_size=(3, 3),
x = self._conv3(x) strides=down_stride,
return (x, y) dilation_rate=dilation_rate,
**dark_conv_args)
self._conv2 = ConvBN(
filters=self._filters // self._filter_scale,
kernel_size=(1, 1),
strides=(1, 1),
**dark_conv_args)
@tf.keras.utils.register_keras_serializable(package="yolo") self._conv3 = ConvBN(
class CSPConnect(tf.keras.layers.Layer): filters=self._filters // self._filter_scale,
"""Sister Layer to the CSPRoute layer. kernel_size=(1, 1),
strides=(1, 1),
**dark_conv_args)
def call(self, inputs, training=None):
if self._downsample:
inputs = self._conv1(inputs)
y = self._conv2(inputs)
x = self._conv3(inputs)
return (x, y)
Merges the partial feature stacks generated by the CSPDownsampling layer,
and the finaly output of the residual stack. Suggested in the CSPNet paper.
@tf.keras.utils.register_keras_serializable(package='yolo')
class CSPConnect(tf.keras.layers.Layer):
"""
Sister Layer to the CSPRoute layer. Merges the partial feature stacks
generated by the CSPDownsampling layer, and the finaly output of the
residual stack. Suggested in the CSPNet paper.
Cross Stage Partial networks (CSPNets) were proposed in: Cross Stage Partial networks (CSPNets) were proposed in:
[1] Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu, Ping-Yang [1] Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu,
Chen, Jun-Wei Hsieh. Ping-Yang Chen, Jun-Wei Hsieh
CSPNet: A New Backbone that can Enhance Learning Capability of CNN. CSPNet: A New Backbone that can Enhance Learning Capability of CNN.
arXiv:1911.11929 arXiv:1911.11929
""" """
def __init__(self, def __init__(self,
filters, filters,
filter_scale=2, filter_scale=2,
activation="mish", drop_final=False,
kernel_initializer="glorot_uniform", drop_first=False,
bias_initializer="zeros", activation='mish',
kernel_regularizer=None, kernel_size=(1, 1),
kernel_initializer='glorot_uniform',
bias_initializer='zeros',
bias_regularizer=None, bias_regularizer=None,
kernel_regularizer=None,
dilation_rate=1,
use_bn=True, use_bn=True,
use_sync_bn=False, use_sync_bn=False,
norm_momentum=0.99, norm_momentum=0.99,
norm_epsilon=0.001, norm_epsilon=0.001,
leaky_alpha=0.1,
**kwargs): **kwargs):
"""Initializes CSPConnect. """
Args: Args:
filters: integer for output depth, or the number of features to learn. filters: integer for output depth, or the number of features to learn
filter_scale: integer dicating (filters//2) or the number of filters in filter_scale: integer dicating (filters//2) or the number of filters in
the partial feature stack. the partial feature stack.
activation: string for activation function to use in layer. activation: string for activation function to use in layer.
...@@ -659,96 +680,112 @@ class CSPConnect(tf.keras.layers.Layer): ...@@ -659,96 +680,112 @@ class CSPConnect(tf.keras.layers.Layer):
bias_regularizer: string to indicate which function to use to regularizer bias_regularizer: string to indicate which function to use to regularizer
bias. bias.
use_bn: boolean for whether to use batch normalization. use_bn: boolean for whether to use batch normalization.
use_sync_bn: boolean for whether sync batch normalization. use_sync_bn: boolean for whether sync batch normalization statistics
norm_momentum: float for moment to use for batch normalization of all batch norm layers to the models global
norm_epsilon: float for batch normalization epsilon statistics (across all input batches).
**kwargs: Keyword Arguments norm_momentum: float for moment to use for batch normalization.
norm_epsilon: float for batch normalization epsilon.
**kwargs: Keyword Arguments.
""" """
super().__init__(**kwargs) super().__init__(**kwargs)
# layer params. # layer params
self._filters = filters self._filters = filters
self._filter_scale = filter_scale self._filter_scale = filter_scale
self._activation = activation self._activation = activation
# Convoultion params. # convoultion params
self._kernel_size = kernel_size
self._kernel_initializer = kernel_initializer self._kernel_initializer = kernel_initializer
self._bias_initializer = bias_initializer self._bias_initializer = bias_initializer
self._kernel_regularizer = kernel_regularizer self._kernel_regularizer = kernel_regularizer
self._bias_regularizer = bias_regularizer self._bias_regularizer = bias_regularizer
self._use_bn = use_bn self._use_bn = use_bn
self._use_sync_bn = use_sync_bn self._use_sync_bn = use_sync_bn
self._norm_moment = norm_momentum self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon self._norm_epsilon = norm_epsilon
self._drop_final = drop_final
self._drop_first = drop_first
self._leaky_alpha = leaky_alpha
def build(self, input_shape): def build(self, input_shape):
self._dark_conv_args = { dark_conv_args = {
"kernel_initializer": self._kernel_initializer, 'kernel_initializer': self._kernel_initializer,
"bias_initializer": self._bias_initializer, 'bias_initializer': self._bias_initializer,
"bias_regularizer": self._bias_regularizer, 'bias_regularizer': self._bias_regularizer,
"use_bn": self._use_bn, 'use_bn': self._use_bn,
"use_sync_bn": self._use_sync_bn, 'use_sync_bn': self._use_sync_bn,
"norm_momentum": self._norm_moment, 'norm_momentum': self._norm_momentum,
"norm_epsilon": self._norm_epsilon, 'norm_epsilon': self._norm_epsilon,
"activation": self._activation, 'activation': self._activation,
"kernel_regularizer": self._kernel_regularizer, 'kernel_regularizer': self._kernel_regularizer,
'leaky_alpha': self._leaky_alpha,
} }
self._conv1 = ConvBN(filters=self._filters // self._filter_scale, if not self._drop_first:
kernel_size=(1, 1), self._conv1 = ConvBN(
strides=(1, 1), filters=self._filters // self._filter_scale,
**self._dark_conv_args) kernel_size=self._kernel_size,
strides=(1, 1),
**dark_conv_args)
self._concat = tf.keras.layers.Concatenate(axis=-1) self._concat = tf.keras.layers.Concatenate(axis=-1)
self._conv2 = ConvBN(filters=self._filters,
kernel_size=(1, 1),
strides=(1, 1),
**self._dark_conv_args)
def call(self, inputs): if not self._drop_final:
self._conv2 = ConvBN(
filters=self._filters,
kernel_size=(1, 1),
strides=(1, 1),
**dark_conv_args)
def call(self, inputs, training=None):
x_prev, x_csp = inputs x_prev, x_csp = inputs
x = self._conv1(x_prev) if not self._drop_first:
x = self._concat([x, x_csp]) x_prev = self._conv1(x_prev)
x = self._conv2(x) x = self._concat([x_prev, x_csp])
# skipped if drop final is true
if not self._drop_final:
x = self._conv2(x)
return x return x
class CSPStack(tf.keras.layers.Layer): class CSPStack(tf.keras.layers.Layer):
"""CSP full stack. """
CSP full stack, combines the route and the connect in case you dont want to
Combines the route and the connect in case you dont want to just quickly wrap jsut quickly wrap an existing callable or list of layers to
an existing callable or list of layers to make it a cross stage partial. make it a cross stage partial. Added for ease of use. you should be able
Added for ease of use. you should be able to wrap any layer stack with a CSP to wrap any layer stack with a CSP independent of wether it belongs
independent of wether it belongs to the Darknet family. if filter_scale = 2, to the Darknet family. if filter_scale = 2, then the blocks in the stack
then the blocks in the stack passed into the the CSP stack should also have passed into the the CSP stack should also have filters = filters/filter_scale
filters = filters/filter_scale.
Cross Stage Partial networks (CSPNets) were proposed in: Cross Stage Partial networks (CSPNets) were proposed in:
[1] Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu, Ping-Yang
Chen, Jun-Wei Hsieh [1] Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu,
Ping-Yang Chen, Jun-Wei Hsieh
CSPNet: A New Backbone that can Enhance Learning Capability of CNN. CSPNet: A New Backbone that can Enhance Learning Capability of CNN.
arXiv:1911.11929 arXiv:1911.11929
""" """
def __init__(self, def __init__(self,
filters, filters,
model_to_wrap=None, model_to_wrap=None,
filter_scale=2, filter_scale=2,
activation="mish", activation='mish',
kernel_initializer="glorot_uniform", kernel_initializer='glorot_uniform',
bias_initializer="zeros", bias_initializer='zeros',
kernel_regularizer=None,
bias_regularizer=None, bias_regularizer=None,
kernel_regularizer=None,
downsample=True, downsample=True,
use_bn=True, use_bn=True,
use_sync_bn=False, use_sync_bn=False,
norm_momentum=0.99, norm_momentum=0.99,
norm_epsilon=0.001, norm_epsilon=0.001,
**kwargs): **kwargs):
"""Initializes CSPStack. """
Args: Args:
filters: integer for output depth, or the number of features to learn.
model_to_wrap: callable Model or a list of callable objects that will model_to_wrap: callable Model or a list of callable objects that will
process the output of CSPRoute, and be input into CSPConnect. List will process the output of CSPRoute, and be input into CSPConnect.
be called sequentially. list will be called sequentially.
downsample: down_sample the input.
filters: integer for output depth, or the number of features to learn.
filter_scale: integer dicating (filters//2) or the number of filters in filter_scale: integer dicating (filters//2) or the number of filters in
the partial feature stack. the partial feature stack.
activation: string for activation function to use in layer. activation: string for activation function to use in layer.
...@@ -760,62 +797,793 @@ class CSPStack(tf.keras.layers.Layer): ...@@ -760,62 +797,793 @@ class CSPStack(tf.keras.layers.Layer):
regularizer weights. regularizer weights.
bias_regularizer: string to indicate which function to use to regularizer bias_regularizer: string to indicate which function to use to regularizer
bias. bias.
downsample: down_sample the input. use_bn: boolean for whether to use batch normalization.
use_bn: boolean for whether to use batch normalization use_sync_bn: boolean for whether sync batch normalization statistics
use_sync_bn: boolean for whether sync batch normalization. of all batch norm layers to the models global statistics
norm_momentum: float for moment to use for batch normalization (across all input batches).
norm_epsilon: float for batch normalization epsilon norm_momentum: float for moment to use for batch normalization.
**kwargs: Keyword Arguments norm_epsilon: float for batch normalization epsilon.
**kwargs: Keyword Arguments.
Raises:
TypeError: model_to_wrap is not a layer or a list of layers
""" """
super().__init__(**kwargs) super().__init__(**kwargs)
# Layer params. # layer params
self._filters = filters self._filters = filters
self._filter_scale = filter_scale self._filter_scale = filter_scale
self._activation = activation self._activation = activation
self._downsample = downsample self._downsample = downsample
# Convoultion params. # convoultion params
self._kernel_initializer = kernel_initializer self._kernel_initializer = kernel_initializer
self._bias_initializer = bias_initializer self._bias_initializer = bias_initializer
self._kernel_regularizer = kernel_regularizer self._kernel_regularizer = kernel_regularizer
self._bias_regularizer = bias_regularizer self._bias_regularizer = bias_regularizer
self._use_bn = use_bn self._use_bn = use_bn
self._use_sync_bn = use_sync_bn self._use_sync_bn = use_sync_bn
self._norm_moment = norm_momentum self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon self._norm_epsilon = norm_epsilon
if model_to_wrap is not None: if model_to_wrap is None:
if isinstance(model_to_wrap, Callable):
self._model_to_wrap = [model_to_wrap]
elif isinstance(model_to_wrap, List):
self._model_to_wrap = model_to_wrap
else:
raise ValueError("The input to the CSPStack must be a list of layers"
"that we can iterate through, or \n a callable")
else:
self._model_to_wrap = [] self._model_to_wrap = []
elif isinstance(model_to_wrap, Callable):
self._model_to_wrap = [model_to_wrap]
elif isinstance(model_to_wrap, List):
self._model_to_wrap = model_to_wrap
else:
raise TypeError(
'the input to the CSPStack must be a list of layers that we can' +
'iterate through, or \n a callable')
def build(self, input_shape): def build(self, input_shape):
self._dark_conv_args = { dark_conv_args = {
"filters": self._filters, 'filters': self._filters,
"filter_scale": self._filter_scale, 'filter_scale': self._filter_scale,
"activation": self._activation, 'activation': self._activation,
"kernel_initializer": self._kernel_initializer, 'kernel_initializer': self._kernel_initializer,
"bias_initializer": self._bias_initializer, 'bias_initializer': self._bias_initializer,
"bias_regularizer": self._bias_regularizer, 'bias_regularizer': self._bias_regularizer,
"use_bn": self._use_bn, 'use_bn': self._use_bn,
"use_sync_bn": self._use_sync_bn, 'use_sync_bn': self._use_sync_bn,
"norm_momentum": self._norm_moment, 'norm_momentum': self._norm_momentum,
"norm_epsilon": self._norm_epsilon, 'norm_epsilon': self._norm_epsilon,
"kernel_regularizer": self._kernel_regularizer, 'kernel_regularizer': self._kernel_regularizer,
} }
self._route = CSPRoute(downsample=self._downsample, **self._dark_conv_args) self._route = CSPRoute(downsample=self._downsample, **dark_conv_args)
self._connect = CSPConnect(**self._dark_conv_args) self._connect = CSPConnect(**dark_conv_args)
return
def call(self, inputs): def call(self, inputs, training=None):
x, x_route = self._route(inputs) x, x_route = self._route(inputs)
for layer in self._model_to_wrap: for layer in self._model_to_wrap:
x = layer(x) x = layer(x)
x = self._connect([x, x_route]) x = self._connect([x, x_route])
return x return x
@tf.keras.utils.register_keras_serializable(package='yolo')
class PathAggregationBlock(tf.keras.layers.Layer):
def __init__(self,
filters=1,
drop_final=True,
kernel_initializer='glorot_uniform',
bias_initializer='zeros',
bias_regularizer=None,
kernel_regularizer=None,
use_bn=True,
use_sync_bn=False,
inverted=False,
norm_momentum=0.99,
norm_epsilon=0.001,
activation='leaky',
leaky_alpha=0.1,
downsample=False,
upsample=False,
upsample_size=2,
**kwargs):
"""
Args:
filters: integer for output depth, or the number of features to learn.
drop_final: do not create the last convolution block.
kernel_initializer: string to indicate which function to use to initialize
weights.
bias_initializer: string to indicate which function to use to initialize
bias.
kernel_regularizer: string to indicate which function to use to
regularizer weights.
bias_regularizer: string to indicate which function to use to regularizer
bias.
use_bn: boolean for whether to use batch normalization.
use_sync_bn: boolean for whether sync batch normalization statistics
of all batch norm layers to the models global statistics
(across all input batches).
inverted: boolean for inverting the order of the convolutions.
norm_momentum: float for moment to use for batch normalization.
norm_epsilon: float for batch normalization epsilon.
activation: string or None for activation function to use in layer,
if None activation is replaced by linear.
leaky_alpha: float to use as alpha if activation function is leaky.
downsample: `bool` for whehter to downwample and merge.
upsample: `bool` for whehter to upsample and merge.
upsample_size: `int` how much to upsample in order to match shapes.
**kwargs: Keyword Arguments.
"""
# darkconv params
self._filters = filters
self._kernel_initializer = kernel_initializer
self._bias_initializer = bias_initializer
self._bias_regularizer = bias_regularizer
self._kernel_regularizer = kernel_regularizer
self._use_bn = use_bn
self._use_sync_bn = use_sync_bn
# normal params
self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon
# activation params
self._conv_activation = activation
self._leaky_alpha = leaky_alpha
self._downsample = downsample
self._upsample = upsample
self._upsample_size = upsample_size
self._drop_final = drop_final
#block params
self._inverted = inverted
super().__init__(**kwargs)
def _build_regular(self, input_shape, kwargs):
if self._downsample:
self._conv = ConvBN(
filters=self._filters,
kernel_size=(3, 3),
strides=(2, 2),
padding='same',
**kwargs)
else:
self._conv = ConvBN(
filters=self._filters,
kernel_size=(1, 1),
strides=(1, 1),
padding='same',
**kwargs)
if not self._drop_final:
self._conv_concat = ConvBN(
filters=self._filters,
kernel_size=(1, 1),
strides=(1, 1),
padding='same',
**kwargs)
def _build_reversed(self, input_shape, kwargs):
if self._downsample:
self._conv_prev = ConvBN(
filters=self._filters,
kernel_size=(3, 3),
strides=(2, 2),
padding='same',
**kwargs)
else:
self._conv_prev = ConvBN(
filters=self._filters,
kernel_size=(1, 1),
strides=(1, 1),
padding='same',
**kwargs)
self._conv_route = ConvBN(
filters=self._filters,
kernel_size=(1, 1),
strides=(1, 1),
padding='same',
**kwargs)
if not self._drop_final:
self._conv_sync = ConvBN(
filters=self._filters,
kernel_size=(1, 1),
strides=(1, 1),
padding='same',
**kwargs)
def build(self, input_shape):
dark_conv_args = {
'kernel_initializer': self._kernel_initializer,
'bias_initializer': self._bias_initializer,
'bias_regularizer': self._bias_regularizer,
'use_bn': self._use_bn,
'use_sync_bn': self._use_sync_bn,
'norm_momentum': self._norm_momentum,
'norm_epsilon': self._norm_epsilon,
'activation': self._conv_activation,
'kernel_regularizer': self._kernel_regularizer,
'leaky_alpha': self._leaky_alpha,
}
if self._inverted:
self._build_reversed(input_shape, dark_conv_args)
else:
self._build_regular(input_shape, dark_conv_args)
self._concat = tf.keras.layers.Concatenate()
super().build(input_shape)
def _call_regular(self, inputs, training=None):
input_to_convolve, input_to_concat = inputs
x_prev = self._conv(input_to_convolve)
if self._upsample:
x_prev = spatial_transform_ops.nearest_upsampling(x_prev,
self._upsample_size)
x = self._concat([x_prev, input_to_concat])
# used in csp conversion
if not self._drop_final:
x = self._conv_concat(x)
return x_prev, x
def _call_reversed(self, inputs, training=None):
x_route, x_prev = inputs
x_prev = self._conv_prev(x_prev)
if self._upsample:
x_prev = spatial_transform_ops.nearest_upsampling(x_prev,
self._upsample_size)
x_route = self._conv_route(x_route)
x = self._concat([x_route, x_prev])
if not self._drop_final:
x = self._conv_sync(x)
return x_prev, x
def call(self, inputs, training=None):
# done this way to prevent confusion in the auto graph
if self._inverted:
return self._call_reversed(inputs, training=training)
else:
return self._call_regular(inputs, training=training)
@tf.keras.utils.register_keras_serializable(package='yolo')
class SPP(tf.keras.layers.Layer):
"""
a non-agregated SPP layer that uses Pooling to gain more performance
"""
def __init__(self, sizes, **kwargs):
self._sizes = list(reversed(sizes))
if len(sizes) == 0:
raise ValueError('More than one maxpool should be specified in SSP block')
super().__init__(**kwargs)
def build(self, input_shape):
maxpools = []
for size in self._sizes:
maxpools.append(
tf.keras.layers.MaxPool2D(
pool_size=(size, size),
strides=(1, 1),
padding='same',
data_format=None))
self._maxpools = maxpools
super().build(input_shape)
def call(self, inputs, training=None):
outputs = []
for maxpool in self._maxpools:
outputs.append(maxpool(inputs))
outputs.append(inputs)
concat_output = tf.keras.layers.concatenate(outputs)
return concat_output
def get_config(self):
layer_config = {'sizes': self._sizes}
layer_config.update(super().get_config())
return layer_config
class SAM(tf.keras.layers.Layer):
"""
[1] Sanghyun Woo, Jongchan Park, Joon-Young Lee, In So Kweon
CBAM: Convolutional Block Attention Module. arXiv:1807.06521
implementation of the Spatial Attention Model (SAM)
"""
def __init__(self,
use_pooling=False,
filter_match=False,
filters=1,
kernel_size=(1, 1),
strides=(1, 1),
padding='same',
dilation_rate=(1, 1),
kernel_initializer='glorot_uniform',
bias_initializer='zeros',
bias_regularizer=None,
kernel_regularizer=None,
use_bn=True,
use_sync_bn=True,
norm_momentum=0.99,
norm_epsilon=0.001,
activation='sigmoid',
output_activation=None,
leaky_alpha=0.1,
**kwargs):
# use_pooling
self._use_pooling = use_pooling
self._filters = filters
self._output_activation = output_activation
self._leaky_alpha = leaky_alpha
self.dark_conv_args = {
'kernel_size': kernel_size,
'strides': strides,
'padding': padding,
'dilation_rate': dilation_rate,
'kernel_initializer': kernel_initializer,
'bias_initializer': bias_initializer,
'bias_regularizer': bias_regularizer,
'use_bn': use_bn,
'use_sync_bn': use_sync_bn,
'norm_momentum': norm_momentum,
'norm_epsilon': norm_epsilon,
'activation': activation,
'kernel_regularizer': kernel_regularizer,
'leaky_alpha': leaky_alpha
}
super().__init__(**kwargs)
def build(self, input_shape):
if self._filters == -1:
self._filters = input_shape[-1]
self._conv = ConvBN(filters=self._filters, **self.dark_conv_args)
if self._output_activation == 'leaky':
self._activation_fn = tf.keras.layers.LeakyReLU(alpha=self._leaky_alpha)
elif self._output_activation == 'mish':
self._activation_fn = lambda x: x * tf.math.tanh(tf.math.softplus(x))
else:
self._activation_fn = tf_utils.get_activation(self._output_activation)
def call(self, inputs, training=None):
if self._use_pooling:
depth_max = tf.reduce_max(inputs, axis=-1, keep_dims=True)
depth_avg = tf.reduce_mean(inputs, axis=-1, keep_dims=True)
input_maps = tf.concat([depth_avg, depth_max], axis=-1)
else:
input_maps = inputs
attention_mask = self._conv(input_maps)
return self._activation_fn(inputs * attention_mask)
class CAM(tf.keras.layers.Layer):
"""
[1] Sanghyun Woo, Jongchan Park, Joon-Young Lee, In So Kweon
CBAM: Convolutional Block Attention Module. arXiv:1807.06521
implementation of the Channel Attention Model (CAM)
"""
def __init__(self,
reduction_ratio=1.0,
kernel_initializer='glorot_uniform',
bias_initializer='zeros',
bias_regularizer=None,
kernel_regularizer=None,
use_bn=False,
use_sync_bn=False,
use_bias=False,
norm_momentum=0.99,
norm_epsilon=0.001,
mlp_activation='linear',
activation='sigmoid',
leaky_alpha=0.1,
**kwargs):
self._reduction_ratio = reduction_ratio
# use_pooling
if use_sync_bn:
self._bn = tf.keras.layers.experimental.SyncBatchNormalization
else:
self._bn = tf.keras.layers.BatchNormalization
if not use_bn:
self._bn = Identity
self._bn_args = {}
else:
self._bn_args = {
'momentum': norm_momentum,
'epsilon': norm_epsilon,
}
self._mlp_args = {
'use_bias': use_bias,
'kernel_initializer': kernel_initializer,
'bias_initializer': bias_initializer,
'bias_regularizer': bias_regularizer,
'activation': mlp_activation,
'kernel_regularizer': kernel_regularizer,
}
self._leaky_alpha = leaky_alpha
self._activation = activation
super().__init__(**kwargs)
def build(self, input_shape):
self._filters = input_shape[-1]
self._mlp = tf.keras.Sequential([
tf.keras.layers.Dense(self._filters, **self._mlp_args),
self._bn(**self._bn_args),
tf.keras.layers.Dense(
int(self._filters * self._reduction_ratio), **self._mlp_args),
self._bn(**self._bn_args),
tf.keras.layers.Dense(self._filters, **self._mlp_args),
self._bn(**self._bn_args),
])
if self._activation == 'leaky':
self._activation_fn = tf.keras.layers.LeakyReLU(alpha=self._leaky_alpha)
elif self._activation == 'mish':
self._activation_fn = lambda x: x * tf.math.tanh(tf.math.softplus(x))
else:
self._activation_fn = tf_utils.get_activation(self._activation)
def call(self, inputs, training=None):
depth_max = self._mlp(tf.reduce_max(inputs, axis=(1, 2)))
depth_avg = self._mlp(tf.reduce_mean(inputs, axis=(1, 2)))
channel_mask = self._activation_fn(depth_avg + depth_max)
channel_mask = tf.expand_dims(channel_mask, axis=1)
attention_mask = tf.expand_dims(channel_mask, axis=1)
return inputs * attention_mask
class CBAM(tf.keras.layers.Layer):
"""
[1] Sanghyun Woo, Jongchan Park, Joon-Young Lee, In So Kweon
CBAM: Convolutional Block Attention Module. arXiv:1807.06521
implementation of the Convolution Block Attention Module (CBAM)
"""
def __init__(self,
use_pooling=False,
filters=1,
reduction_ratio=1.0,
kernel_size=(1, 1),
strides=(1, 1),
padding='same',
dilation_rate=(1, 1),
kernel_initializer='glorot_uniform',
bias_initializer='zeros',
bias_regularizer=None,
kernel_regularizer=None,
use_bn=True,
use_sync_bn=False,
norm_momentum=0.99,
norm_epsilon=0.001,
mlp_activation=None,
activation='sigmoid',
leaky_alpha=0.1,
**kwargs):
# use_pooling
self._sam_args = {
'use_pooling': use_pooling,
'filters': filters,
'kernel_size': kernel_size,
'strides': strides,
'padding': padding,
'dilation_rate': dilation_rate,
}
self._cam_args = {
'reduction_ratio': reduction_ratio,
'mlp_activation': mlp_activation
}
self._common_args = {
'kernel_initializer': kernel_initializer,
'bias_initializer': bias_initializer,
'bias_regularizer': bias_regularizer,
'use_bn': use_bn,
'use_sync_bn': use_sync_bn,
'norm_momentum': norm_momentum,
'norm_epsilon': norm_epsilon,
'activation': activation,
'kernel_regularizer': kernel_regularizer,
'leaky_alpha': leaky_alpha
}
self._cam_args.update(self._common_args)
self._sam_args.update(self._common_args)
super().__init__(**kwargs)
def build(self, input_shape):
self._cam = CAM(**self._cam_args)
self._sam = SAM(**self._sam_args)
def call(self, inputs, training=None):
return self._sam(self._cam(inputs))
@tf.keras.utils.register_keras_serializable(package='yolo')
class DarkRouteProcess(tf.keras.layers.Layer):
"""
process darknet outputs and connect back bone to head more generalizably
Abstracts repetition of DarkConv objects that is common in YOLO.
It is used like the following:
x = ConvBN(1024, (3, 3), (1, 1))(x)
proc = DarkRouteProcess(filters = 1024,
repetitions = 3,
insert_spp = False)(x)
"""
def __init__(
self,
filters=2,
repetitions=2,
insert_spp=False,
insert_sam=False,
insert_cbam=False,
csp_stack=0,
csp_scale=2,
kernel_initializer='glorot_uniform',
bias_initializer='zeros',
bias_regularizer=None,
use_sync_bn=False,
kernel_regularizer=None, # default find where is it is stated
norm_momentum=0.99,
norm_epsilon=0.001,
block_invert=False,
activation='leaky',
leaky_alpha=0.1,
spp_keys=None,
**kwargs):
"""
Args:
filters: the number of filters to be used in all subsequent layers
filters should be the depth of the tensor input into this layer,
as no downsampling can be done within this layer object.
repetitions: number of times to repeat the processign nodes
for tiny: 1 repition, no spp allowed
for spp: insert_spp = True, and allow for 3+ repetitions
for regular: insert_spp = False, and allow for 3+ repetitions.
insert_spp: bool if true add the spatial pyramid pooling layer.
kernel_initializer: method to use to initializa kernel weights.
bias_initializer: method to use to initialize the bias of the conv
layers.
norm_momentum: batch norm parameter see TensorFlow documentation.
norm_epsilon: batch norm parameter see TensorFlow documentation.
activation: activation function to use in processing.
leaky_alpha: if leaky acitivation function, the alpha to use in
processing the relu input.
Returns:
callable tensorflow layer
Raises:
None
"""
super().__init__(**kwargs)
# darkconv params
self._filters = filters
self._use_sync_bn = use_sync_bn
self._kernel_initializer = kernel_initializer
self._bias_initializer = bias_initializer
self._bias_regularizer = bias_regularizer
self._kernel_regularizer = kernel_regularizer
# normal params
self._norm_momentum = norm_momentum
self._norm_epsilon = norm_epsilon
# activation params
self._activation = activation
self._leaky_alpha = leaky_alpha
repetitions += (2 * int(insert_spp))
if repetitions == 1:
block_invert = True
self._repetitions = repetitions
self.layer_list, self.outputs = self._get_base_layers()
if csp_stack > 0:
self._csp_scale = csp_scale
csp_stack += (2 * int(insert_spp))
self._csp_filters = lambda x: x // csp_scale
self._convert_csp(self.layer_list, self.outputs, csp_stack)
block_invert = False
self._csp_stack = csp_stack
if block_invert:
self._conv1_filters = lambda x: x
self._conv2_filters = lambda x: x // 2
self._conv1_kernel = (3, 3)
self._conv2_kernel = (1, 1)
else:
self._conv1_filters = lambda x: x // 2
self._conv2_filters = lambda x: x
self._conv1_kernel = (1, 1)
self._conv2_kernel = (3, 3)
# insert SPP will always add to the total nuber of layer, never replace
if insert_spp:
self._spp_keys = spp_keys if spp_keys is not None else [5, 9, 13]
self.layer_list = self._insert_spp(self.layer_list)
if repetitions > 1:
self.outputs[-2] = True
if insert_sam:
self.layer_list = self._insert_sam(self.layer_list, self.outputs)
self._repetitions += 1
self.outputs[-1] = True
def _get_base_layers(self):
layer_list = []
outputs = []
for i in range(self._repetitions):
layers = ['conv1'] * ((i + 1) % 2) + ['conv2'] * (i % 2)
layer_list.extend(layers)
outputs = [False] + outputs
return layer_list, outputs
def _insert_spp(self, layer_list):
if len(layer_list) <= 3:
layer_list[1] = 'spp'
else:
layer_list[3] = 'spp'
return layer_list
def _convert_csp(self, layer_list, outputs, csp_stack_size):
layer_list[0] = 'csp_route'
layer_list.insert(csp_stack_size - 1, 'csp_connect')
outputs.insert(csp_stack_size - 1, False)
return layer_list, outputs
def _insert_sam(self, layer_list, outputs):
if len(layer_list) >= 2 and layer_list[-2] != 'spp':
layer_list.insert(-2, 'sam')
outputs.insert(-1, True)
else:
layer_list.insert(-1, 'sam')
outputs.insert(-1, False)
return layer_list
def _conv1(self, filters, kwargs, csp=False):
if csp:
filters_ = self._csp_filters
else:
filters_ = self._conv1_filters
x1 = ConvBN(
filters=filters_(filters),
kernel_size=self._conv1_kernel,
strides=(1, 1),
padding='same',
use_bn=True,
**kwargs)
return x1
def _conv2(self, filters, kwargs, csp=False):
if csp:
filters_ = self._csp_filters
else:
filters_ = self._conv2_filters
x1 = ConvBN(
filters=filters_(filters),
kernel_size=self._conv2_kernel,
strides=(1, 1),
padding='same',
use_bn=True,
**kwargs)
return x1
def _csp_route(self, filters, kwargs):
x1 = CSPRoute(
filters=filters,
filter_scale=self._csp_scale,
downsample=False,
**kwargs)
return x1
def _csp_connect(self, filters, kwargs):
x1 = CSPConnect(filters=filters, drop_final=True, drop_first=True, **kwargs)
return x1
def _spp(self, filters, kwargs):
x1 = SPP(self._spp_keys)
return x1
def _sam(self, filters, kwargs):
x1 = SAM(filters=-1, use_pooling=False, use_bn=True, **kwargs)
return x1
def build(self, input_shape):
dark_conv_args = {
'activation': self._activation,
'kernel_initializer': self._kernel_initializer,
'bias_initializer': self._bias_initializer,
'bias_regularizer': self._bias_regularizer,
'use_sync_bn': self._use_sync_bn,
'norm_momentum': self._norm_momentum,
'norm_epsilon': self._norm_epsilon,
'kernel_regularizer': self._kernel_regularizer,
'leaky_alpha': self._leaky_alpha,
}
csp = False
self.layers = []
for layer in self.layer_list:
if layer == 'csp_route':
self.layers.append(self._csp_route(self._filters, dark_conv_args))
csp = True
elif layer == 'csp_connect':
self.layers.append(self._csp_connect(self._filters, dark_conv_args))
csp = False
elif layer == 'conv1':
self.layers.append(self._conv1(self._filters, dark_conv_args, csp=csp))
elif layer == 'conv2':
self.layers.append(self._conv2(self._filters, dark_conv_args, csp=csp))
elif layer == 'spp':
self.layers.append(self._spp(self._filters, dark_conv_args))
elif layer == 'sam':
self.layers.append(self._sam(-1, _args))
self._lim = len(self.layers)
super().build(input_shape)
def _call_regular(self, inputs, training=None):
# check efficiency
x = inputs
x_prev = x
output_prev = True
for i, (layer, output) in enumerate(zip(self.layers, self.outputs)):
if output_prev:
x_prev = x
x = layer(x)
output_prev = output
return x_prev, x
def _call_csp(self, inputs, training=None):
# check efficiency
x = inputs
x_prev = x
output_prev = True
x_route = None
for i, (layer, output) in enumerate(zip(self.layers, self.outputs)):
if output_prev:
x_prev = x
if i == 0:
x, x_route = layer(x)
elif i == self._csp_stack - 1:
x = layer([x, x_route])
else:
x = layer(x)
output_prev = output
return x_prev, x
def call(self, inputs, training=None):
if self._csp_stack > 0:
return self._call_csp(inputs, training=training)
else:
return self._call_regular(inputs)
official/vision/beta/projects/yolo/modeling/layers/nn_blocks_test.py
View file @ 63719f08
...@@ -13,18 +13,17 @@ ...@@ -13,18 +13,17 @@
# limitations under the License. # limitations under the License.
# Lint as: python3 # Lint as: python3
from absl.testing import parameterized
import numpy as np
import tensorflow as tf import tensorflow as tf
import numpy as np
from absl.testing import parameterized
from official.vision.beta.projects.yolo.modeling.layers import nn_blocks from official.vision.beta.projects.yolo.modeling.layers import nn_blocks
class CSPConnectTest(tf.test.TestCase, parameterized.TestCase): class CSPConnectTest(tf.test.TestCase, parameterized.TestCase):
@parameterized.named_parameters(("same", 224, 224, 64, 1), @parameterized.named_parameters(('same', 224, 224, 64, 1),
("downsample", 224, 224, 64, 2)) ('downsample', 224, 224, 64, 2))
def test_pass_through(self, width, height, filters, mod): def test_pass_through(self, width, height, filters, mod):
x = tf.keras.Input(shape=(width, height, filters)) x = tf.keras.Input(shape=(width, height, filters))
test_layer = nn_blocks.CSPRoute(filters=filters, filter_scale=mod) test_layer = nn_blocks.CSPRoute(filters=filters, filter_scale=mod)
...@@ -38,8 +37,8 @@ class CSPConnectTest(tf.test.TestCase, parameterized.TestCase): ...@@ -38,8 +37,8 @@ class CSPConnectTest(tf.test.TestCase, parameterized.TestCase):
[None, np.ceil(width // 2), [None, np.ceil(width // 2),
np.ceil(height // 2), (filters)]) np.ceil(height // 2), (filters)])
@parameterized.named_parameters(("same", 224, 224, 64, 1), @parameterized.named_parameters(('same', 224, 224, 64, 1),
("downsample", 224, 224, 128, 2)) ('downsample', 224, 224, 128, 2))
def test_gradient_pass_though(self, filters, width, height, mod): def test_gradient_pass_though(self, filters, width, height, mod):
loss = tf.keras.losses.MeanSquaredError() loss = tf.keras.losses.MeanSquaredError()
optimizer = tf.keras.optimizers.SGD() optimizer = tf.keras.optimizers.SGD()
...@@ -49,10 +48,11 @@ class CSPConnectTest(tf.test.TestCase, parameterized.TestCase): ...@@ -49,10 +48,11 @@ class CSPConnectTest(tf.test.TestCase, parameterized.TestCase):
init = tf.random_normal_initializer() init = tf.random_normal_initializer()
x = tf.Variable( x = tf.Variable(
initial_value=init(shape=(1, width, height, filters), dtype=tf.float32)) initial_value=init(shape=(1, width, height, filters), dtype=tf.float32))
y = tf.Variable(initial_value=init(shape=(1, int(np.ceil(width // 2)), y = tf.Variable(
int(np.ceil(height // 2)), initial_value=init(
filters), shape=(1, int(np.ceil(width // 2)), int(np.ceil(height // 2)),
dtype=tf.float32)) filters),
dtype=tf.float32))
with tf.GradientTape() as tape: with tf.GradientTape() as tape:
x_hat, x_prev = test_layer(x) x_hat, x_prev = test_layer(x)
...@@ -66,12 +66,12 @@ class CSPConnectTest(tf.test.TestCase, parameterized.TestCase): ...@@ -66,12 +66,12 @@ class CSPConnectTest(tf.test.TestCase, parameterized.TestCase):
class CSPRouteTest(tf.test.TestCase, parameterized.TestCase): class CSPRouteTest(tf.test.TestCase, parameterized.TestCase):
@parameterized.named_parameters(("same", 224, 224, 64, 1), @parameterized.named_parameters(('same', 224, 224, 64, 1),
("downsample", 224, 224, 64, 2)) ('downsample', 224, 224, 64, 2))
def test_pass_through(self, width, height, filters, mod): def test_pass_through(self, width, height, filters, mod):
x = tf.keras.Input(shape=(width, height, filters)) x = tf.keras.Input(shape=(width, height, filters))
test_layer = nn_blocks.CSPRoute(filters=filters, filter_scale=mod) test_layer = nn_blocks.CSPRoute(filters=filters, filter_scale=mod)
outx, _ = test_layer(x) outx, px = test_layer(x)
print(outx) print(outx)
print(outx.shape.as_list()) print(outx.shape.as_list())
self.assertAllEqual( self.assertAllEqual(
...@@ -79,8 +79,8 @@ class CSPRouteTest(tf.test.TestCase, parameterized.TestCase): ...@@ -79,8 +79,8 @@ class CSPRouteTest(tf.test.TestCase, parameterized.TestCase):
[None, np.ceil(width // 2), [None, np.ceil(width // 2),
np.ceil(height // 2), (filters / mod)]) np.ceil(height // 2), (filters / mod)])
@parameterized.named_parameters(("same", 224, 224, 64, 1), @parameterized.named_parameters(('same', 224, 224, 64, 1),
("downsample", 224, 224, 128, 2)) ('downsample', 224, 224, 128, 2))
def test_gradient_pass_though(self, filters, width, height, mod): def test_gradient_pass_though(self, filters, width, height, mod):
loss = tf.keras.losses.MeanSquaredError() loss = tf.keras.losses.MeanSquaredError()
optimizer = tf.keras.optimizers.SGD() optimizer = tf.keras.optimizers.SGD()
...@@ -90,10 +90,11 @@ class CSPRouteTest(tf.test.TestCase, parameterized.TestCase): ...@@ -90,10 +90,11 @@ class CSPRouteTest(tf.test.TestCase, parameterized.TestCase):
init = tf.random_normal_initializer() init = tf.random_normal_initializer()
x = tf.Variable( x = tf.Variable(
initial_value=init(shape=(1, width, height, filters), dtype=tf.float32)) initial_value=init(shape=(1, width, height, filters), dtype=tf.float32))
y = tf.Variable(initial_value=init(shape=(1, int(np.ceil(width // 2)), y = tf.Variable(
int(np.ceil(height // 2)), initial_value=init(
filters), shape=(1, int(np.ceil(width // 2)), int(np.ceil(height // 2)),
dtype=tf.float32)) filters),
dtype=tf.float32))
with tf.GradientTape() as tape: with tf.GradientTape() as tape:
x_hat, x_prev = test_layer(x) x_hat, x_prev = test_layer(x)
...@@ -107,11 +108,11 @@ class CSPRouteTest(tf.test.TestCase, parameterized.TestCase): ...@@ -107,11 +108,11 @@ class CSPRouteTest(tf.test.TestCase, parameterized.TestCase):
class CSPStackTest(tf.test.TestCase, parameterized.TestCase): class CSPStackTest(tf.test.TestCase, parameterized.TestCase):
def build_layer( def build_layer(self, layer_type, filters, filter_scale, count, stack_type,
self, layer_type, filters, filter_scale, count, stack_type, downsample): downsample):
if stack_type is not None: if stack_type is not None:
layers = [] layers = []
if layer_type == "residual": if layer_type == 'residual':
for _ in range(count): for _ in range(count):
layers.append( layers.append(
nn_blocks.DarkResidual( nn_blocks.DarkResidual(
...@@ -120,7 +121,7 @@ class CSPStackTest(tf.test.TestCase, parameterized.TestCase): ...@@ -120,7 +121,7 @@ class CSPStackTest(tf.test.TestCase, parameterized.TestCase):
for _ in range(count): for _ in range(count):
layers.append(nn_blocks.ConvBN(filters=filters)) layers.append(nn_blocks.ConvBN(filters=filters))
if stack_type == "model": if stack_type == 'model':
layers = tf.keras.Sequential(layers=layers) layers = tf.keras.Sequential(layers=layers)
else: else:
layers = None layers = None
...@@ -133,10 +134,10 @@ class CSPStackTest(tf.test.TestCase, parameterized.TestCase): ...@@ -133,10 +134,10 @@ class CSPStackTest(tf.test.TestCase, parameterized.TestCase):
return stack return stack
@parameterized.named_parameters( @parameterized.named_parameters(
("no_stack", 224, 224, 64, 2, "residual", None, 0, True), ('no_stack', 224, 224, 64, 2, 'residual', None, 0, True),
("residual_stack", 224, 224, 64, 2, "residual", "list", 2, True), ('residual_stack', 224, 224, 64, 2, 'residual', 'list', 2, True),
("conv_stack", 224, 224, 64, 2, "conv", "list", 3, False), ('conv_stack', 224, 224, 64, 2, 'conv', 'list', 3, False),
("callable_no_scale", 224, 224, 64, 1, "residual", "model", 5, False)) ('callable_no_scale', 224, 224, 64, 1, 'residual', 'model', 5, False))
def test_pass_through(self, width, height, filters, mod, layer_type, def test_pass_through(self, width, height, filters, mod, layer_type,
stack_type, count, downsample): stack_type, count, downsample):
x = tf.keras.Input(shape=(width, height, filters)) x = tf.keras.Input(shape=(width, height, filters))
...@@ -152,10 +153,10 @@ class CSPStackTest(tf.test.TestCase, parameterized.TestCase): ...@@ -152,10 +153,10 @@ class CSPStackTest(tf.test.TestCase, parameterized.TestCase):
self.assertAllEqual(outx.shape.as_list(), [None, width, height, filters]) self.assertAllEqual(outx.shape.as_list(), [None, width, height, filters])
@parameterized.named_parameters( @parameterized.named_parameters(
("no_stack", 224, 224, 64, 2, "residual", None, 0, True), ('no_stack', 224, 224, 64, 2, 'residual', None, 0, True),
("residual_stack", 224, 224, 64, 2, "residual", "list", 2, True), ('residual_stack', 224, 224, 64, 2, 'residual', 'list', 2, True),
("conv_stack", 224, 224, 64, 2, "conv", "list", 3, False), ('conv_stack', 224, 224, 64, 2, 'conv', 'list', 3, False),
("callable_no_scale", 224, 224, 64, 1, "residual", "model", 5, False)) ('callable_no_scale', 224, 224, 64, 1, 'residual', 'model', 5, False))
def test_gradient_pass_though(self, width, height, filters, mod, layer_type, def test_gradient_pass_though(self, width, height, filters, mod, layer_type,
stack_type, count, downsample): stack_type, count, downsample):
loss = tf.keras.losses.MeanSquaredError() loss = tf.keras.losses.MeanSquaredError()
...@@ -188,10 +189,10 @@ class CSPStackTest(tf.test.TestCase, parameterized.TestCase): ...@@ -188,10 +189,10 @@ class CSPStackTest(tf.test.TestCase, parameterized.TestCase):
class ConvBNTest(tf.test.TestCase, parameterized.TestCase): class ConvBNTest(tf.test.TestCase, parameterized.TestCase):
@parameterized.named_parameters( @parameterized.named_parameters(
("valid", (3, 3), "valid", (1, 1)), ("same", (3, 3), "same", (1, 1)), ('valid', (3, 3), 'valid', (1, 1)), ('same', (3, 3), 'same', (1, 1)),
("downsample", (3, 3), "same", (2, 2)), ("test", (1, 1), "valid", (1, 1))) ('downsample', (3, 3), 'same', (2, 2)), ('test', (1, 1), 'valid', (1, 1)))
def test_pass_through(self, kernel_size, padding, strides): def test_pass_through(self, kernel_size, padding, strides):
if padding == "same": if padding == 'same':
pad_const = 1 pad_const = 1
else: else:
pad_const = 0 pad_const = 0
...@@ -212,16 +213,16 @@ class ConvBNTest(tf.test.TestCase, parameterized.TestCase): ...@@ -212,16 +213,16 @@ class ConvBNTest(tf.test.TestCase, parameterized.TestCase):
print(test) print(test)
self.assertAllEqual(outx.shape.as_list(), test) self.assertAllEqual(outx.shape.as_list(), test)
@parameterized.named_parameters(("filters", 3)) @parameterized.named_parameters(('filters', 3))
def test_gradient_pass_though(self, filters): def test_gradient_pass_though(self, filters):
loss = tf.keras.losses.MeanSquaredError() loss = tf.keras.losses.MeanSquaredError()
optimizer = tf.keras.optimizers.SGD() optimizer = tf.keras.optimizers.SGD()
with tf.device("/CPU:0"): with tf.device('/CPU:0'):
test_layer = nn_blocks.ConvBN(filters, kernel_size=(3, 3), padding="same") test_layer = nn_blocks.ConvBN(filters, kernel_size=(3, 3), padding='same')
init = tf.random_normal_initializer() init = tf.random_normal_initializer()
x = tf.Variable(initial_value=init(shape=(1, 224, 224, x = tf.Variable(
3), dtype=tf.float32)) initial_value=init(shape=(1, 224, 224, 3), dtype=tf.float32))
y = tf.Variable( y = tf.Variable(
initial_value=init(shape=(1, 224, 224, filters), dtype=tf.float32)) initial_value=init(shape=(1, 224, 224, filters), dtype=tf.float32))
...@@ -235,9 +236,9 @@ class ConvBNTest(tf.test.TestCase, parameterized.TestCase): ...@@ -235,9 +236,9 @@ class ConvBNTest(tf.test.TestCase, parameterized.TestCase):
class DarkResidualTest(tf.test.TestCase, parameterized.TestCase): class DarkResidualTest(tf.test.TestCase, parameterized.TestCase):
@parameterized.named_parameters(("same", 224, 224, 64, False), @parameterized.named_parameters(('same', 224, 224, 64, False),
("downsample", 223, 223, 32, True), ('downsample', 223, 223, 32, True),
("oddball", 223, 223, 32, False)) ('oddball', 223, 223, 32, False))
def test_pass_through(self, width, height, filters, downsample): def test_pass_through(self, width, height, filters, downsample):
mod = 1 mod = 1
if downsample: if downsample:
...@@ -252,9 +253,9 @@ class DarkResidualTest(tf.test.TestCase, parameterized.TestCase): ...@@ -252,9 +253,9 @@ class DarkResidualTest(tf.test.TestCase, parameterized.TestCase):
[None, np.ceil(width / mod), [None, np.ceil(width / mod),
np.ceil(height / mod), filters]) np.ceil(height / mod), filters])
@parameterized.named_parameters(("same", 64, 224, 224, False), @parameterized.named_parameters(('same', 64, 224, 224, False),
("downsample", 32, 223, 223, True), ('downsample', 32, 223, 223, True),
("oddball", 32, 223, 223, False)) ('oddball', 32, 223, 223, False))
def test_gradient_pass_though(self, filters, width, height, downsample): def test_gradient_pass_though(self, filters, width, height, downsample):
loss = tf.keras.losses.MeanSquaredError() loss = tf.keras.losses.MeanSquaredError()
optimizer = tf.keras.optimizers.SGD() optimizer = tf.keras.optimizers.SGD()
...@@ -268,10 +269,11 @@ class DarkResidualTest(tf.test.TestCase, parameterized.TestCase): ...@@ -268,10 +269,11 @@ class DarkResidualTest(tf.test.TestCase, parameterized.TestCase):
init = tf.random_normal_initializer() init = tf.random_normal_initializer()
x = tf.Variable( x = tf.Variable(
initial_value=init(shape=(1, width, height, filters), dtype=tf.float32)) initial_value=init(shape=(1, width, height, filters), dtype=tf.float32))
y = tf.Variable(initial_value=init(shape=(1, int(np.ceil(width / mod)), y = tf.Variable(
int(np.ceil(height / mod)), initial_value=init(
filters), shape=(1, int(np.ceil(width / mod)), int(np.ceil(height / mod)),
dtype=tf.float32)) filters),
dtype=tf.float32))
with tf.GradientTape() as tape: with tf.GradientTape() as tape:
x_hat = test_layer(x) x_hat = test_layer(x)
...@@ -281,5 +283,102 @@ class DarkResidualTest(tf.test.TestCase, parameterized.TestCase): ...@@ -281,5 +283,102 @@ class DarkResidualTest(tf.test.TestCase, parameterized.TestCase):
self.assertNotIn(None, grad) self.assertNotIn(None, grad)
if __name__ == "__main__":
class DarkSppTest(tf.test.TestCase, parameterized.TestCase):
@parameterized.named_parameters(('RouteProcessSpp', 224, 224, 3, [5, 9, 13]),
('test1', 300, 300, 10, [2, 3, 4, 5]),
('test2', 256, 256, 5, [10]))
def test_pass_through(self, width, height, channels, sizes):
x = tf.keras.Input(shape=(width, height, channels))
test_layer = nn_blocks.SPP(sizes=sizes)
outx = test_layer(x)
self.assertAllEqual(outx.shape.as_list(),
[None, width, height, channels * (len(sizes) + 1)])
return
@parameterized.named_parameters(('RouteProcessSpp', 224, 224, 3, [5, 9, 13]),
('test1', 300, 300, 10, [2, 3, 4, 5]),
('test2', 256, 256, 5, [10]))
def test_gradient_pass_though(self, width, height, channels, sizes):
loss = tf.keras.losses.MeanSquaredError()
optimizer = tf.keras.optimizers.SGD()
test_layer = nn_blocks.SPP(sizes=sizes)
init = tf.random_normal_initializer()
x = tf.Variable(
initial_value=init(
shape=(1, width, height, channels), dtype=tf.float32))
y = tf.Variable(
initial_value=init(
shape=(1, width, height, channels * (len(sizes) + 1)),
dtype=tf.float32))
with tf.GradientTape() as tape:
x_hat = test_layer(x)
grad_loss = loss(x_hat, y)
grad = tape.gradient(grad_loss, test_layer.trainable_variables)
optimizer.apply_gradients(zip(grad, test_layer.trainable_variables))
self.assertNotIn(None, grad)
return
class DarkRouteProcessTest(tf.test.TestCase, parameterized.TestCase):
@parameterized.named_parameters(
('test1', 224, 224, 64, 7, False), ('test2', 223, 223, 32, 3, False),
('tiny', 223, 223, 16, 1, False), ('spp', 224, 224, 64, 7, False))
def test_pass_through(self, width, height, filters, repetitions, spp):
x = tf.keras.Input(shape=(width, height, filters))
test_layer = nn_blocks.DarkRouteProcess(
filters=filters, repetitions=repetitions, insert_spp=spp)
outx = test_layer(x)
self.assertEqual(len(outx), 2, msg='len(outx) != 2')
if repetitions == 1:
filter_y1 = filters
else:
filter_y1 = filters // 2
self.assertAllEqual(outx[1].shape.as_list(), [None, width, height, filter_y1])
self.assertAllEqual(
filters % 2,
0,
msg='Output of a DarkRouteProcess layer has an odd number of filters')
self.assertAllEqual(outx[0].shape.as_list(), [None, width, height, filters])
@parameterized.named_parameters(
('test1', 224, 224, 64, 7, False), ('test2', 223, 223, 32, 3, False),
('tiny', 223, 223, 16, 1, False), ('spp', 224, 224, 64, 7, False))
def test_gradient_pass_though(self, width, height, filters, repetitions, spp):
loss = tf.keras.losses.MeanSquaredError()
optimizer = tf.keras.optimizers.SGD()
test_layer = nn_blocks.DarkRouteProcess(
filters=filters, repetitions=repetitions, insert_spp=spp)
if repetitions == 1:
filter_y1 = filters
else:
filter_y1 = filters // 2
init = tf.random_normal_initializer()
x = tf.Variable(
initial_value=init(shape=(1, width, height, filters), dtype=tf.float32))
y_0 = tf.Variable(
initial_value=init(shape=(1, width, height, filters), dtype=tf.float32))
y_1 = tf.Variable(
initial_value=init(shape=(1, width, height, filter_y1), dtype=tf.float32))
with tf.GradientTape() as tape:
x_hat_0, x_hat_1 = test_layer(x)
grad_loss_0 = loss(x_hat_0, y_0)
grad_loss_1 = loss(x_hat_1, y_1)
grad = tape.gradient([grad_loss_0, grad_loss_1],
test_layer.trainable_variables)
optimizer.apply_gradients(zip(grad, test_layer.trainable_variables))
self.assertNotIn(None, grad)
return
if __name__ == '__main__':
tf.test.main() tf.test.main()
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment