config_util.py

# Copyright 2017 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.
# ==============================================================================
"""Functions for reading and updating configuration files."""

import os
import tensorflow as tf

from google.protobuf import text_format

from tensorflow.python.lib.io import file_io

from object_detection.protos import eval_pb2
from object_detection.protos import graph_rewriter_pb2
from object_detection.protos import input_reader_pb2
from object_detection.protos import model_pb2
from object_detection.protos import pipeline_pb2
from object_detection.protos import train_pb2


def get_image_resizer_config(model_config):
  """Returns the image resizer config from a model config.

  Args:
    model_config: A model_pb2.DetectionModel.

  Returns:
    An image_resizer_pb2.ImageResizer.

  Raises:
    ValueError: If the model type is not recognized.
  """
  meta_architecture = model_config.WhichOneof("model")
  if meta_architecture == "faster_rcnn":
    return model_config.faster_rcnn.image_resizer
  if meta_architecture == "ssd":
    return model_config.ssd.image_resizer

  raise ValueError("Unknown model type: {}".format(meta_architecture))


def get_spatial_image_size(image_resizer_config):
  """Returns expected spatial size of the output image from a given config.

  Args:
    image_resizer_config: An image_resizer_pb2.ImageResizer.

  Returns:
    A list of two integers of the form [height, width]. `height` and `width` are
    set  -1 if they cannot be determined during graph construction.

  Raises:
    ValueError: If the model type is not recognized.
  """
  if image_resizer_config.HasField("fixed_shape_resizer"):
    return [
        image_resizer_config.fixed_shape_resizer.height,
        image_resizer_config.fixed_shape_resizer.width
    ]
  if image_resizer_config.HasField("keep_aspect_ratio_resizer"):
    if image_resizer_config.keep_aspect_ratio_resizer.pad_to_max_dimension:
      return [image_resizer_config.keep_aspect_ratio_resizer.max_dimension] * 2
    else:
      return [-1, -1]
  raise ValueError("Unknown image resizer type.")


def get_configs_from_pipeline_file(pipeline_config_path):
  """Reads config from a file containing pipeline_pb2.TrainEvalPipelineConfig.

  Args:
    pipeline_config_path: Path to pipeline_pb2.TrainEvalPipelineConfig text
      proto.

  Returns:
    Dictionary of configuration objects. Keys are `model`, `train_config`,
      `train_input_config`, `eval_config`, `eval_input_config`. Value are the
      corresponding config objects.
  """
  pipeline_config = pipeline_pb2.TrainEvalPipelineConfig()
  with tf.gfile.GFile(pipeline_config_path, "r") as f:
    proto_str = f.read()
    text_format.Merge(proto_str, pipeline_config)
  return create_configs_from_pipeline_proto(pipeline_config)


def create_configs_from_pipeline_proto(pipeline_config):
  """Creates a configs dictionary from pipeline_pb2.TrainEvalPipelineConfig.

  Args:
    pipeline_config: pipeline_pb2.TrainEvalPipelineConfig proto object.

  Returns:
    Dictionary of configuration objects. Keys are `model`, `train_config`,
      `train_input_config`, `eval_config`, `eval_input_config`. Value are the
      corresponding config objects.
  """
  configs = {}
  configs["model"] = pipeline_config.model
  configs["train_config"] = pipeline_config.train_config
  configs["train_input_config"] = pipeline_config.train_input_reader
  configs["eval_config"] = pipeline_config.eval_config
  configs["eval_input_config"] = pipeline_config.eval_input_reader
  if pipeline_config.HasField("graph_rewriter"):
    configs["graph_rewriter_config"] = pipeline_config.graph_rewriter

  return configs


def get_graph_rewriter_config_from_file(graph_rewriter_config_file):
  """Parses config for graph rewriter.

  Args:
    graph_rewriter_config_file: file path to the graph rewriter config.

  Returns:
    graph_rewriter_pb2.GraphRewriter proto
  """
  graph_rewriter_config = graph_rewriter_pb2.GraphRewriter()
  with tf.gfile.GFile(graph_rewriter_config_file, "r") as f:
    text_format.Merge(f.read(), graph_rewriter_config)
  return graph_rewriter_config


def create_pipeline_proto_from_configs(configs):
  """Creates a pipeline_pb2.TrainEvalPipelineConfig from configs dictionary.

  This function performs the inverse operation of
  create_configs_from_pipeline_proto().

  Args:
    configs: Dictionary of configs. See get_configs_from_pipeline_file().

  Returns:
    A fully populated pipeline_pb2.TrainEvalPipelineConfig.
  """
  pipeline_config = pipeline_pb2.TrainEvalPipelineConfig()
  pipeline_config.model.CopyFrom(configs["model"])
  pipeline_config.train_config.CopyFrom(configs["train_config"])
  pipeline_config.train_input_reader.CopyFrom(configs["train_input_config"])
  pipeline_config.eval_config.CopyFrom(configs["eval_config"])
  pipeline_config.eval_input_reader.CopyFrom(configs["eval_input_config"])
  if "graph_rewriter_config" in configs:
    pipeline_config.graph_rewriter.CopyFrom(configs["graph_rewriter_config"])
  return pipeline_config


def save_pipeline_config(pipeline_config, directory):
  """Saves a pipeline config text file to disk.

  Args:
    pipeline_config: A pipeline_pb2.TrainEvalPipelineConfig.
    directory: The model directory into which the pipeline config file will be
      saved.
  """
  if not file_io.file_exists(directory):
    file_io.recursive_create_dir(directory)
  pipeline_config_path = os.path.join(directory, "pipeline.config")
  config_text = text_format.MessageToString(pipeline_config)
  with tf.gfile.Open(pipeline_config_path, "wb") as f:
    tf.logging.info("Writing pipeline config file to %s",
                    pipeline_config_path)
    f.write(config_text)


def get_configs_from_multiple_files(model_config_path="",
                                    train_config_path="",
                                    train_input_config_path="",
                                    eval_config_path="",
                                    eval_input_config_path="",
                                    graph_rewriter_config_path=""):
  """Reads training configuration from multiple config files.

  Args:
    model_config_path: Path to model_pb2.DetectionModel.
    train_config_path: Path to train_pb2.TrainConfig.
    train_input_config_path: Path to input_reader_pb2.InputReader.
    eval_config_path: Path to eval_pb2.EvalConfig.
    eval_input_config_path: Path to input_reader_pb2.InputReader.
    graph_rewriter_config_path: Path to graph_rewriter_pb2.GraphRewriter.

  Returns:
    Dictionary of configuration objects. Keys are `model`, `train_config`,
      `train_input_config`, `eval_config`, `eval_input_config`. Key/Values are
        returned only for valid (non-empty) strings.
  """
  configs = {}
  if model_config_path:
    model_config = model_pb2.DetectionModel()
    with tf.gfile.GFile(model_config_path, "r") as f:
      text_format.Merge(f.read(), model_config)
      configs["model"] = model_config

  if train_config_path:
    train_config = train_pb2.TrainConfig()
    with tf.gfile.GFile(train_config_path, "r") as f:
      text_format.Merge(f.read(), train_config)
      configs["train_config"] = train_config

  if train_input_config_path:
    train_input_config = input_reader_pb2.InputReader()
    with tf.gfile.GFile(train_input_config_path, "r") as f:
      text_format.Merge(f.read(), train_input_config)
      configs["train_input_config"] = train_input_config

  if eval_config_path:
    eval_config = eval_pb2.EvalConfig()
    with tf.gfile.GFile(eval_config_path, "r") as f:
      text_format.Merge(f.read(), eval_config)
      configs["eval_config"] = eval_config

  if eval_input_config_path:
    eval_input_config = input_reader_pb2.InputReader()
    with tf.gfile.GFile(eval_input_config_path, "r") as f:
      text_format.Merge(f.read(), eval_input_config)
      configs["eval_input_config"] = eval_input_config

  if graph_rewriter_config_path:
    configs["graph_rewriter_config"] = get_graph_rewriter_config_from_file(
        graph_rewriter_config_path)

  return configs


def get_number_of_classes(model_config):
  """Returns the number of classes for a detection model.

  Args:
    model_config: A model_pb2.DetectionModel.

  Returns:
    Number of classes.

  Raises:
    ValueError: If the model type is not recognized.
  """
  meta_architecture = model_config.WhichOneof("model")
  if meta_architecture == "faster_rcnn":
    return model_config.faster_rcnn.num_classes
  if meta_architecture == "ssd":
    return model_config.ssd.num_classes

  raise ValueError("Expected the model to be one of 'faster_rcnn' or 'ssd'.")


def get_optimizer_type(train_config):
  """Returns the optimizer type for training.

  Args:
    train_config: A train_pb2.TrainConfig.

  Returns:
    The type of the optimizer
  """
  return train_config.optimizer.WhichOneof("optimizer")


def get_learning_rate_type(optimizer_config):
  """Returns the learning rate type for training.

  Args:
    optimizer_config: An optimizer_pb2.Optimizer.

  Returns:
    The type of the learning rate.
  """
  return optimizer_config.learning_rate.WhichOneof("learning_rate")


def _is_generic_key(key):
  """Determines whether the key starts with a generic config dictionary key."""
  for prefix in [
      "graph_rewriter_config",
      "model",
      "train_input_config",
      "train_input_config",
      "train_config"]:
    if key.startswith(prefix + "."):
      return True
  return False


def merge_external_params_with_configs(configs, hparams=None, **kwargs):
  """Updates `configs` dictionary based on supplied parameters.

  This utility is for modifying specific fields in the object detection configs.
  Say that one would like to experiment with different learning rates, momentum
  values, or batch sizes. Rather than creating a new config text file for each
  experiment, one can use a single base config file, and update particular
  values.

  There are two types of field overrides:
  1. Strategy-based overrides, which update multiple relevant configuration
  options. For example, updating `learning_rate` will update both the warmup and
  final learning rates.
  2. Generic key/value, which update a specific parameter based on namespaced
  configuration keys. For example,
  `model.ssd.loss.hard_example_miner.max_negatives_per_positive` will update the
  hard example miner configuration for an SSD model config. Generic overrides
  are automatically detected based on the namespaced keys.

  Args:
    configs: Dictionary of configuration objects. See outputs from
      get_configs_from_pipeline_file() or get_configs_from_multiple_files().
    hparams: A `HParams`.
    **kwargs: Extra keyword arguments that are treated the same way as
      attribute/value pairs in `hparams`. Note that hyperparameters with the
      same names will override keyword arguments.

  Returns:
    `configs` dictionary.
  """

  if hparams:
    kwargs.update(hparams.values())
  for key, value in kwargs.items():
    tf.logging.info("Maybe overwriting %s: %s", key, value)
    # pylint: disable=g-explicit-bool-comparison
    if value == "" or value is None:
      continue
    # pylint: enable=g-explicit-bool-comparison
    if key == "learning_rate":
      _update_initial_learning_rate(configs, value)
    elif key == "batch_size":
      _update_batch_size(configs, value)
    elif key == "momentum_optimizer_value":
      _update_momentum_optimizer_value(configs, value)
    elif key == "classification_localization_weight_ratio":
      # Localization weight is fixed to 1.0.
      _update_classification_localization_weight_ratio(configs, value)
    elif key == "focal_loss_gamma":
      _update_focal_loss_gamma(configs, value)
    elif key == "focal_loss_alpha":
      _update_focal_loss_alpha(configs, value)
    elif key == "train_steps":
      _update_train_steps(configs, value)
    elif key == "eval_steps":
      _update_eval_steps(configs, value)
    elif key == "train_input_path":
      _update_input_path(configs["train_input_config"], value)
    elif key == "eval_input_path":
      _update_input_path(configs["eval_input_config"], value)
    elif key == "label_map_path":
      _update_label_map_path(configs, value)
    elif key == "mask_type":
      _update_mask_type(configs, value)
    elif key == "eval_with_moving_averages":
      _update_use_moving_averages(configs, value)
    elif _is_generic_key(key):
      _update_generic(configs, key, value)
    else:
      tf.logging.info("Ignoring config override key: %s", key)
  return configs


def _update_initial_learning_rate(configs, learning_rate):
  """Updates `configs` to reflect the new initial learning rate.

  This function updates the initial learning rate. For learning rate schedules,
  all other defined learning rates in the pipeline config are scaled to maintain
  their same ratio with the initial learning rate.
  The configs dictionary is updated in place, and hence not returned.

  Args:
    configs: Dictionary of configuration objects. See outputs from
      get_configs_from_pipeline_file() or get_configs_from_multiple_files().
    learning_rate: Initial learning rate for optimizer.

  Raises:
    TypeError: if optimizer type is not supported, or if learning rate type is
      not supported.
  """

  optimizer_type = get_optimizer_type(configs["train_config"])
  if optimizer_type == "rms_prop_optimizer":
    optimizer_config = configs["train_config"].optimizer.rms_prop_optimizer
  elif optimizer_type == "momentum_optimizer":
    optimizer_config = configs["train_config"].optimizer.momentum_optimizer
  elif optimizer_type == "adam_optimizer":
    optimizer_config = configs["train_config"].optimizer.adam_optimizer
  else:
    raise TypeError("Optimizer %s is not supported." % optimizer_type)

  learning_rate_type = get_learning_rate_type(optimizer_config)
  if learning_rate_type == "constant_learning_rate":
    constant_lr = optimizer_config.learning_rate.constant_learning_rate
    constant_lr.learning_rate = learning_rate
  elif learning_rate_type == "exponential_decay_learning_rate":
    exponential_lr = (
        optimizer_config.learning_rate.exponential_decay_learning_rate)
    exponential_lr.initial_learning_rate = learning_rate
  elif learning_rate_type == "manual_step_learning_rate":
    manual_lr = optimizer_config.learning_rate.manual_step_learning_rate
    original_learning_rate = manual_lr.initial_learning_rate
    learning_rate_scaling = float(learning_rate) / original_learning_rate
    manual_lr.initial_learning_rate = learning_rate
    for schedule in manual_lr.schedule:
      schedule.learning_rate *= learning_rate_scaling
  elif learning_rate_type == "cosine_decay_learning_rate":
    cosine_lr = optimizer_config.learning_rate.cosine_decay_learning_rate
    learning_rate_base = cosine_lr.learning_rate_base
    warmup_learning_rate = cosine_lr.warmup_learning_rate
    warmup_scale_factor = warmup_learning_rate / learning_rate_base
    cosine_lr.learning_rate_base = learning_rate
    cosine_lr.warmup_learning_rate = warmup_scale_factor * learning_rate
  else:
    raise TypeError("Learning rate %s is not supported." % learning_rate_type)


def _update_batch_size(configs, batch_size):
  """Updates `configs` to reflect the new training batch size.

  The configs dictionary is updated in place, and hence not returned.

  Args:
    configs: Dictionary of configuration objects. See outputs from
      get_configs_from_pipeline_file() or get_configs_from_multiple_files().
    batch_size: Batch size to use for training (Ideally a power of 2). Inputs
      are rounded, and capped to be 1 or greater.
  """
  configs["train_config"].batch_size = max(1, int(round(batch_size)))


def _validate_message_has_field(message, field):
  if not message.HasField(field):
    raise ValueError("Expecting message to have field %s" % field)


def _update_generic(configs, key, value):
  """Update a pipeline configuration parameter based on a generic key/value.

  Args:
    configs: Dictionary of pipeline configuration protos.
    key: A string key, dot-delimited to represent the argument key.
      e.g. "model.ssd.train_config.batch_size"
    value: A value to set the argument to. The type of the value must match the
      type for the protocol buffer. Note that setting the wrong type will
      result in a TypeError.
      e.g. 42

  Raises:
    ValueError if the message key does not match the existing proto fields.
    TypeError the value type doesn't match the protobuf field type.
  """
  fields = key.split(".")
  first_field = fields.pop(0)
  last_field = fields.pop()
  message = configs[first_field]
  for field in fields:
    _validate_message_has_field(message, field)
    message = getattr(message, field)
  _validate_message_has_field(message, last_field)
  setattr(message, last_field, value)


def _update_momentum_optimizer_value(configs, momentum):
  """Updates `configs` to reflect the new momentum value.

  Momentum is only supported for RMSPropOptimizer and MomentumOptimizer. For any
  other optimizer, no changes take place. The configs dictionary is updated in
  place, and hence not returned.

  Args:
    configs: Dictionary of configuration objects. See outputs from
      get_configs_from_pipeline_file() or get_configs_from_multiple_files().
    momentum: New momentum value. Values are clipped at 0.0 and 1.0.

  Raises:
    TypeError: If the optimizer type is not `rms_prop_optimizer` or
    `momentum_optimizer`.
  """
  optimizer_type = get_optimizer_type(configs["train_config"])
  if optimizer_type == "rms_prop_optimizer":
    optimizer_config = configs["train_config"].optimizer.rms_prop_optimizer
  elif optimizer_type == "momentum_optimizer":
    optimizer_config = configs["train_config"].optimizer.momentum_optimizer
  else:
    raise TypeError("Optimizer type must be one of `rms_prop_optimizer` or "
                    "`momentum_optimizer`.")

  optimizer_config.momentum_optimizer_value = min(max(0.0, momentum), 1.0)


def _update_classification_localization_weight_ratio(configs, ratio):
  """Updates the classification/localization weight loss ratio.

  Detection models usually define a loss weight for both classification and
  objectness. This function updates the weights such that the ratio between
  classification weight to localization weight is the ratio provided.
  Arbitrarily, localization weight is set to 1.0.

  Note that in the case of Faster R-CNN, this same ratio is applied to the first
  stage objectness loss weight relative to localization loss weight.

  The configs dictionary is updated in place, and hence not returned.

  Args:
    configs: Dictionary of configuration objects. See outputs from
      get_configs_from_pipeline_file() or get_configs_from_multiple_files().
    ratio: Desired ratio of classification (and/or objectness) loss weight to
      localization loss weight.
  """
  meta_architecture = configs["model"].WhichOneof("model")
  if meta_architecture == "faster_rcnn":
    model = configs["model"].faster_rcnn
    model.first_stage_localization_loss_weight = 1.0
    model.first_stage_objectness_loss_weight = ratio
    model.second_stage_localization_loss_weight = 1.0
    model.second_stage_classification_loss_weight = ratio
  if meta_architecture == "ssd":
    model = configs["model"].ssd
    model.loss.localization_weight = 1.0
    model.loss.classification_weight = ratio


def _get_classification_loss(model_config):
  """Returns the classification loss for a model."""
  meta_architecture = model_config.WhichOneof("model")
  if meta_architecture == "faster_rcnn":
    model = model_config.faster_rcnn
    classification_loss = model.second_stage_classification_loss
  elif meta_architecture == "ssd":
    model = model_config.ssd
    classification_loss = model.loss.classification_loss
  else:
    raise TypeError("Did not recognize the model architecture.")
  return classification_loss


def _update_focal_loss_gamma(configs, gamma):
  """Updates the gamma value for a sigmoid focal loss.

  The configs dictionary is updated in place, and hence not returned.

  Args:
    configs: Dictionary of configuration objects. See outputs from
      get_configs_from_pipeline_file() or get_configs_from_multiple_files().
    gamma: Exponent term in focal loss.

  Raises:
    TypeError: If the classification loss is not `weighted_sigmoid_focal`.
  """
  classification_loss = _get_classification_loss(configs["model"])
  classification_loss_type = classification_loss.WhichOneof(
      "classification_loss")
  if classification_loss_type != "weighted_sigmoid_focal":
    raise TypeError("Classification loss must be `weighted_sigmoid_focal`.")
  classification_loss.weighted_sigmoid_focal.gamma = gamma


def _update_focal_loss_alpha(configs, alpha):
  """Updates the alpha value for a sigmoid focal loss.

  The configs dictionary is updated in place, and hence not returned.

  Args:
    configs: Dictionary of configuration objects. See outputs from
      get_configs_from_pipeline_file() or get_configs_from_multiple_files().
    alpha: Class weight multiplier for sigmoid loss.

  Raises:
    TypeError: If the classification loss is not `weighted_sigmoid_focal`.
  """
  classification_loss = _get_classification_loss(configs["model"])
  classification_loss_type = classification_loss.WhichOneof(
      "classification_loss")
  if classification_loss_type != "weighted_sigmoid_focal":
    raise TypeError("Classification loss must be `weighted_sigmoid_focal`.")
  classification_loss.weighted_sigmoid_focal.alpha = alpha


def _update_train_steps(configs, train_steps):
  """Updates `configs` to reflect new number of training steps."""
  configs["train_config"].num_steps = int(train_steps)


def _update_eval_steps(configs, eval_steps):
  """Updates `configs` to reflect new number of eval steps per evaluation."""
  configs["eval_config"].num_examples = int(eval_steps)


def _update_input_path(input_config, input_path):
  """Updates input configuration to reflect a new input path.

  The input_config object is updated in place, and hence not returned.

  Args:
    input_config: A input_reader_pb2.InputReader.
    input_path: A path to data or list of paths.

  Raises:
    TypeError: if input reader type is not `tf_record_input_reader`.
  """
  input_reader_type = input_config.WhichOneof("input_reader")
  if input_reader_type == "tf_record_input_reader":
    input_config.tf_record_input_reader.ClearField("input_path")
    if isinstance(input_path, list):
      input_config.tf_record_input_reader.input_path.extend(input_path)
    else:
      input_config.tf_record_input_reader.input_path.append(input_path)
  else:
    raise TypeError("Input reader type must be `tf_record_input_reader`.")


def _update_label_map_path(configs, label_map_path):
  """Updates the label map path for both train and eval input readers.

  The configs dictionary is updated in place, and hence not returned.

  Args:
    configs: Dictionary of configuration objects. See outputs from
      get_configs_from_pipeline_file() or get_configs_from_multiple_files().
    label_map_path: New path to `StringIntLabelMap` pbtxt file.
  """
  configs["train_input_config"].label_map_path = label_map_path
  configs["eval_input_config"].label_map_path = label_map_path


def _update_mask_type(configs, mask_type):
  """Updates the mask type for both train and eval input readers.

  The configs dictionary is updated in place, and hence not returned.

  Args:
    configs: Dictionary of configuration objects. See outputs from
      get_configs_from_pipeline_file() or get_configs_from_multiple_files().
    mask_type: A string name representing a value of
      input_reader_pb2.InstanceMaskType
  """
  configs["train_input_config"].mask_type = mask_type
  configs["eval_input_config"].mask_type = mask_type


def _update_use_moving_averages(configs, use_moving_averages):
  """Updates the eval config option to use or not use moving averages.

  The configs dictionary is updated in place, and hence not returned.

  Args:
    configs: Dictionary of configuration objects. See outputs from
      get_configs_from_pipeline_file() or get_configs_from_multiple_files().
    use_moving_averages: Boolean indicating whether moving average variables
      should be loaded during evaluation.
  """
  configs["eval_config"].use_moving_averages = use_moving_averages