Skip to content

GitLab

Commit 4684103f authored by Leif's avatar Leif
Browse files

Merge remote-tracking branch 'PaddlePaddle/dygraph' into dygraph

parents 2bea3736 19204e8f
Hide whitespace changes
Inline Side-by-side
Showing with 799 additions and 168 deletions
deploy/pdserving/config.yml
View file @ 4684103f
#rpc端口, rpc_port和http_port不允许同时为空。当rpc_port为空且http_port不为空时，会自动将rpc_port设置为http_port+1 #rpc端口, rpc_port和http_port不允许同时为空。当rpc_port为空且http_port不为空时，会自动将rpc_port设置为http_port+1
rpc_port: 18090 rpc_port: 18091
#http端口, rpc_port和http_port不允许同时为空。当rpc_port可用且http_port为空时，不自动生成http_port #http端口, rpc_port和http_port不允许同时为空。当rpc_port可用且http_port为空时，不自动生成http_port
http_port: 9999 http_port: 9998
#worker_num, 最大并发数。当build_dag_each_worker=True时, 框架会创建worker_num个进程，每个进程内构建grpcSever和DAG #worker_num, 最大并发数。当build_dag_each_worker=True时, 框架会创建worker_num个进程，每个进程内构建grpcSever和DAG
##当build_dag_each_worker=False时，框架会设置主线程grpc线程池的max_workers=worker_num ##当build_dag_each_worker=False时，框架会设置主线程grpc线程池的max_workers=worker_num
worker_num: 20 worker_num: 10
#build_dag_each_worker, False，框架在进程内创建一条DAG；True，框架会每个进程内创建多个独立的DAG #build_dag_each_worker, False，框架在进程内创建一条DAG；True，框架会每个进程内创建多个独立的DAG
build_dag_each_worker: false build_dag_each_worker: False
dag: dag:
#op资源类型, True, 为线程模型；False，为进程模型 #op资源类型, True, 为线程模型；False，为进程模型
is_thread_op: False is_thread_op: False
#重试次数 #重试次数
retry: 1 retry: 10
#使用性能分析, True，生成Timeline性能数据，对性能有一定影响；False为不使用 #使用性能分析, True，生成Timeline性能数据，对性能有一定影响；False为不使用
use_profile: False use_profile: True
tracer: tracer:
interval_s: 10 interval_s: 10
op: op:
det: det:
#并发数，is_thread_op=True时，为线程并发；否则为进程并发 #并发数，is_thread_op=True时，为线程并发；否则为进程并发
concurrency: 4 concurrency: 8
#当op配置没有server_endpoints时，从local_service_conf读取本地服务配置 #当op配置没有server_endpoints时，从local_service_conf读取本地服务配置
local_service_conf: local_service_conf:
...@@ -34,18 +34,18 @@ op: ...@@ -34,18 +34,18 @@ op:
client_type: local_predictor client_type: local_predictor
#det模型路径 #det模型路径
model_config: /paddle/serving/models/det_serving_server/ #ocr_det_model model_config: ./ppocr_det_mobile_2.0_serving
#Fetch结果列表，以client_config中fetch_var的alias_name为准 #Fetch结果列表，以client_config中fetch_var的alias_name为准
fetch_list: ["save_infer_model/scale_0.tmp_1"] fetch_list: ["save_infer_model/scale_0.tmp_1"]
#计算硬件ID，当devices为""或不写时为CPU预测；当devices为"0", "0,1,2"时为GPU预测，表示使用的GPU卡 #计算硬件ID，当devices为""或不写时为CPU预测；当devices为"0", "0,1,2"时为GPU预测，表示使用的GPU卡
devices: "2" devices: "0"
ir_optim: True ir_optim: True
rec: rec:
#并发数，is_thread_op=True时，为线程并发；否则为进程并发 #并发数，is_thread_op=True时，为线程并发；否则为进程并发
concurrency: 1 concurrency: 4
#超时时间, 单位ms #超时时间, 单位ms
timeout: -1 timeout: -1
...@@ -60,12 +60,12 @@ op: ...@@ -60,12 +60,12 @@ op:
client_type: local_predictor client_type: local_predictor
#rec模型路径 #rec模型路径
model_config: /paddle/serving/models/rec_serving_server/ #ocr_rec_model model_config: ./ppocr_rec_mobile_2.0_serving
#Fetch结果列表，以client_config中fetch_var的alias_name为准 #Fetch结果列表，以client_config中fetch_var的alias_name为准
fetch_list: ["save_infer_model/scale_0.tmp_1"] #["ctc_greedy_decoder_0.tmp_0", "softmax_0.tmp_0"] fetch_list: ["save_infer_model/scale_0.tmp_1"]
#计算硬件ID，当devices为""或不写时为CPU预测；当devices为"0", "0,1,2"时为GPU预测，表示使用的GPU卡 #计算硬件ID，当devices为""或不写时为CPU预测；当devices为"0", "0,1,2"时为GPU预测，表示使用的GPU卡
devices: "2" devices: "0"
ir_optim: True ir_optim: True
deploy/pdserving/ocr_reader.py
View file @ 4684103f
...@@ -21,7 +21,6 @@ import sys ...@@ -21,7 +21,6 @@ import sys
import argparse import argparse
import string import string
from copy import deepcopy from copy import deepcopy
import paddle
class DetResizeForTest(object): class DetResizeForTest(object):
...@@ -34,12 +33,12 @@ class DetResizeForTest(object): ...@@ -34,12 +33,12 @@ class DetResizeForTest(object):
elif 'limit_side_len' in kwargs: elif 'limit_side_len' in kwargs:
self.limit_side_len = kwargs['limit_side_len'] self.limit_side_len = kwargs['limit_side_len']
self.limit_type = kwargs.get('limit_type', 'min') self.limit_type = kwargs.get('limit_type', 'min')
elif 'resize_long' in kwargs: elif 'resize_short' in kwargs:
self.resize_type = 2
self.resize_long = kwargs.get('resize_long', 960)
else:
self.limit_side_len = 736 self.limit_side_len = 736
self.limit_type = 'min' self.limit_type = 'min'
else:
self.resize_type = 2
self.resize_long = kwargs.get('resize_long', 960)
def __call__(self, data): def __call__(self, data):
img = deepcopy(data) img = deepcopy(data)
...@@ -227,8 +226,6 @@ class CTCLabelDecode(BaseRecLabelDecode): ...@@ -227,8 +226,6 @@ class CTCLabelDecode(BaseRecLabelDecode):
super(CTCLabelDecode, self).__init__(config) super(CTCLabelDecode, self).__init__(config)
def __call__(self, preds, label=None, *args, **kwargs): def __call__(self, preds, label=None, *args, **kwargs):
if isinstance(preds, paddle.Tensor):
preds = preds.numpy()
preds_idx = preds.argmax(axis=2) preds_idx = preds.argmax(axis=2)
preds_prob = preds.max(axis=2) preds_prob = preds.max(axis=2)
text = self.decode(preds_idx, preds_prob, is_remove_duplicate=True) text = self.decode(preds_idx, preds_prob, is_remove_duplicate=True)
......
deploy/pdserving/pipeline_http_client.py
View file @ 4684103f
...@@ -23,8 +23,8 @@ def cv2_to_base64(image): ...@@ -23,8 +23,8 @@ def cv2_to_base64(image):
return base64.b64encode(image).decode('utf8') return base64.b64encode(image).decode('utf8')
url = "http://127.0.0.1:9999/ocr/prediction" url = "http://127.0.0.1:9998/ocr/prediction"
test_img_dir = "../doc/imgs/" test_img_dir = "../../doc/imgs/"
for idx, img_file in enumerate(os.listdir(test_img_dir)): for idx, img_file in enumerate(os.listdir(test_img_dir)):
with open(os.path.join(test_img_dir, img_file), 'rb') as file: with open(os.path.join(test_img_dir, img_file), 'rb') as file:
image_data1 = file.read() image_data1 = file.read()
...@@ -36,5 +36,5 @@ for idx, img_file in enumerate(os.listdir(test_img_dir)): ...@@ -36,5 +36,5 @@ for idx, img_file in enumerate(os.listdir(test_img_dir)):
r = requests.post(url=url, data=json.dumps(data)) r = requests.post(url=url, data=json.dumps(data))
print(r.json()) print(r.json())
test_img_dir = "../doc/imgs/" test_img_dir = "../../doc/imgs/"
print("==> total number of test imgs: ", len(os.listdir(test_img_dir))) print("==> total number of test imgs: ", len(os.listdir(test_img_dir)))
deploy/pdserving/pipeline_rpc_client.py
View file @ 4684103f
...@@ -23,7 +23,7 @@ import base64 ...@@ -23,7 +23,7 @@ import base64
import os import os
client = PipelineClient() client = PipelineClient()
client.connect(['127.0.0.1:18090']) client.connect(['127.0.0.1:18091'])
def cv2_to_base64(image): def cv2_to_base64(image):
...@@ -39,4 +39,3 @@ for img_file in os.listdir(test_img_dir): ...@@ -39,4 +39,3 @@ for img_file in os.listdir(test_img_dir):
for i in range(1): for i in range(1):
ret = client.predict(feed_dict={"image": image}, fetch=["res"]) ret = client.predict(feed_dict={"image": image}, fetch=["res"])
print(ret) print(ret)
#print(ret)
deploy/pdserving/web_service.py
View file @ 4684103f
...@@ -11,10 +11,7 @@ ...@@ -11,10 +11,7 @@
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and # See the License for the specific language governing permissions and
# limitations under the License. # limitations under the License.
try: from paddle_serving_server.web_service import WebService, Op
from paddle_serving_server_gpu.web_service import WebService, Op
except ImportError:
from paddle_serving_server.web_service import WebService, Op
import logging import logging
import numpy as np import numpy as np
...@@ -48,28 +45,24 @@ class DetOp(Op): ...@@ -48,28 +45,24 @@ class DetOp(Op):
def preprocess(self, input_dicts, data_id, log_id): def preprocess(self, input_dicts, data_id, log_id):
(_, input_dict), = input_dicts.items() (_, input_dict), = input_dicts.items()
data = base64.b64decode(input_dict["image"].encode('utf8')) data = base64.b64decode(input_dict["image"].encode('utf8'))
self.raw_im = data
data = np.fromstring(data, np.uint8) data = np.fromstring(data, np.uint8)
# Note: class variables(self.var) can only be used in process op mode # Note: class variables(self.var) can only be used in process op mode
im = cv2.imdecode(data, cv2.IMREAD_COLOR) im = cv2.imdecode(data, cv2.IMREAD_COLOR)
self.im = im
self.ori_h, self.ori_w, _ = im.shape self.ori_h, self.ori_w, _ = im.shape
det_img = self.det_preprocess(im)
det_img = self.det_preprocess(self.im)
_, self.new_h, self.new_w = det_img.shape _, self.new_h, self.new_w = det_img.shape
print("det image shape", det_img.shape)
return {"x": det_img[np.newaxis, :].copy()}, False, None, "" return {"x": det_img[np.newaxis, :].copy()}, False, None, ""
def postprocess(self, input_dicts, fetch_dict, log_id): def postprocess(self, input_dicts, fetch_dict, log_id):
print("input_dicts: ", input_dicts)
det_out = fetch_dict["save_infer_model/scale_0.tmp_1"] det_out = fetch_dict["save_infer_model/scale_0.tmp_1"]
ratio_list = [ ratio_list = [
float(self.new_h) / self.ori_h, float(self.new_w) / self.ori_w float(self.new_h) / self.ori_h, float(self.new_w) / self.ori_w
] ]
dt_boxes_list = self.post_func(det_out, [ratio_list]) dt_boxes_list = self.post_func(det_out, [ratio_list])
dt_boxes = self.filter_func(dt_boxes_list[0], [self.ori_h, self.ori_w]) dt_boxes = self.filter_func(dt_boxes_list[0], [self.ori_h, self.ori_w])
out_dict = {"dt_boxes": dt_boxes, "image": self.im} out_dict = {"dt_boxes": dt_boxes, "image": self.raw_im}
print("out dict", out_dict["dt_boxes"])
return out_dict, None, "" return out_dict, None, ""
...@@ -83,35 +76,75 @@ class RecOp(Op): ...@@ -83,35 +76,75 @@ class RecOp(Op):
def preprocess(self, input_dicts, data_id, log_id): def preprocess(self, input_dicts, data_id, log_id):
(_, input_dict), = input_dicts.items() (_, input_dict), = input_dicts.items()
im = input_dict["image"] raw_im = input_dict["image"]
data = np.frombuffer(raw_im, np.uint8)
im = cv2.imdecode(data, cv2.IMREAD_COLOR)
dt_boxes = input_dict["dt_boxes"] dt_boxes = input_dict["dt_boxes"]
dt_boxes = self.sorted_boxes(dt_boxes) dt_boxes = self.sorted_boxes(dt_boxes)
feed_list = [] feed_list = []
img_list = [] img_list = []
max_wh_ratio = 0 max_wh_ratio = 0
for i, dtbox in enumerate(dt_boxes): ## Many mini-batchs, the type of feed_data is list.
boximg = self.get_rotate_crop_image(im, dt_boxes[i]) max_batch_size = 6 # len(dt_boxes)
img_list.append(boximg)
h, w = boximg.shape[0:2] # If max_batch_size is 0, skipping predict stage
wh_ratio = w * 1.0 / h if max_batch_size == 0:
max_wh_ratio = max(max_wh_ratio, wh_ratio) return {}, True, None, ""
_, w, h = self.ocr_reader.resize_norm_img(img_list[0], boxes_size = len(dt_boxes)
max_wh_ratio).shape batch_size = boxes_size // max_batch_size
rem = boxes_size % max_batch_size
imgs = np.zeros((len(img_list), 3, w, h)).astype('float32') for bt_idx in range(0, batch_size + 1):
for id, img in enumerate(img_list): imgs = None
norm_img = self.ocr_reader.resize_norm_img(img, max_wh_ratio) boxes_num_in_one_batch = 0
imgs[id] = norm_img if bt_idx == batch_size:
print("rec image shape", imgs.shape) if rem == 0:
feed = {"x": imgs.copy()} continue
return feed, False, None, "" else:
boxes_num_in_one_batch = rem
def postprocess(self, input_dicts, fetch_dict, log_id): elif bt_idx < batch_size:
rec_res = self.ocr_reader.postprocess(fetch_dict, with_score=True) boxes_num_in_one_batch = max_batch_size
res_lst = [] else:
for res in rec_res: _LOGGER.error("batch_size error, bt_idx={}, batch_size={}".
res_lst.append(res[0]) format(bt_idx, batch_size))
res = {"res": str(res_lst)} break
start = bt_idx * max_batch_size
end = start + boxes_num_in_one_batch
img_list = []
for box_idx in range(start, end):
boximg = self.get_rotate_crop_image(im, dt_boxes[box_idx])
img_list.append(boximg)
h, w = boximg.shape[0:2]
wh_ratio = w * 1.0 / h
max_wh_ratio = max(max_wh_ratio, wh_ratio)
_, w, h = self.ocr_reader.resize_norm_img(img_list[0],
max_wh_ratio).shape
imgs = np.zeros((boxes_num_in_one_batch, 3, w, h)).astype('float32')
for id, img in enumerate(img_list):
norm_img = self.ocr_reader.resize_norm_img(img, max_wh_ratio)
imgs[id] = norm_img
feed = {"x": imgs.copy()}
feed_list.append(feed)
return feed_list, False, None, ""
def postprocess(self, input_dicts, fetch_data, log_id):
res_list = []
if isinstance(fetch_data, dict):
if len(fetch_data) > 0:
rec_batch_res = self.ocr_reader.postprocess(
fetch_data, with_score=True)
for res in rec_batch_res:
res_list.append(res[0])
elif isinstance(fetch_data, list):
for one_batch in fetch_data:
one_batch_res = self.ocr_reader.postprocess(
one_batch, with_score=True)
for res in one_batch_res:
res_list.append(res[0])
res = {"res": str(res_list)}
return res, None, "" return res, None, ""
......
deploy/pdserving/win/ocr_reader.py 0 → 100644
View file @ 4684103f
# Copyright (c) 2021 PaddlePaddle 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.
import cv2
import copy
import numpy as np
import math
import re
import sys
import argparse
import string
from copy import deepcopy
class DetResizeForTest(object):
def __init__(self, **kwargs):
super(DetResizeForTest, self).__init__()
self.resize_type = 0
if 'image_shape' in kwargs:
self.image_shape = kwargs['image_shape']
self.resize_type = 1
elif 'limit_side_len' in kwargs:
self.limit_side_len = kwargs['limit_side_len']
self.limit_type = kwargs.get('limit_type', 'min')
elif 'resize_short' in kwargs:
self.limit_side_len = 736
self.limit_type = 'min'
else:
self.resize_type = 2
self.resize_long = kwargs.get('resize_long', 960)
def __call__(self, data):
img = deepcopy(data)
src_h, src_w, _ = img.shape
if self.resize_type == 0:
img, [ratio_h, ratio_w] = self.resize_image_type0(img)
elif self.resize_type == 2:
img, [ratio_h, ratio_w] = self.resize_image_type2(img)
else:
img, [ratio_h, ratio_w] = self.resize_image_type1(img)
return img
def resize_image_type1(self, img):
resize_h, resize_w = self.image_shape
ori_h, ori_w = img.shape[:2] # (h, w, c)
ratio_h = float(resize_h) / ori_h
ratio_w = float(resize_w) / ori_w
img = cv2.resize(img, (int(resize_w), int(resize_h)))
return img, [ratio_h, ratio_w]
def resize_image_type0(self, img):
"""
resize image to a size multiple of 32 which is required by the network
args:
img(array): array with shape [h, w, c]
return(tuple):
img, (ratio_h, ratio_w)
"""
limit_side_len = self.limit_side_len
h, w, _ = img.shape
# limit the max side
if self.limit_type == 'max':
if max(h, w) > limit_side_len:
if h > w:
ratio = float(limit_side_len) / h
else:
ratio = float(limit_side_len) / w
else:
ratio = 1.
else:
if min(h, w) < limit_side_len:
if h < w:
ratio = float(limit_side_len) / h
else:
ratio = float(limit_side_len) / w
else:
ratio = 1.
resize_h = int(h * ratio)
resize_w = int(w * ratio)
resize_h = int(round(resize_h / 32) * 32)
resize_w = int(round(resize_w / 32) * 32)
try:
if int(resize_w) <= 0 or int(resize_h) <= 0:
return None, (None, None)
img = cv2.resize(img, (int(resize_w), int(resize_h)))
except:
print(img.shape, resize_w, resize_h)
sys.exit(0)
ratio_h = resize_h / float(h)
ratio_w = resize_w / float(w)
# return img, np.array([h, w])
return img, [ratio_h, ratio_w]
def resize_image_type2(self, img):
h, w, _ = img.shape
resize_w = w
resize_h = h
# Fix the longer side
if resize_h > resize_w:
ratio = float(self.resize_long) / resize_h
else:
ratio = float(self.resize_long) / resize_w
resize_h = int(resize_h * ratio)
resize_w = int(resize_w * ratio)
max_stride = 128
resize_h = (resize_h + max_stride - 1) // max_stride * max_stride
resize_w = (resize_w + max_stride - 1) // max_stride * max_stride
img = cv2.resize(img, (int(resize_w), int(resize_h)))
ratio_h = resize_h / float(h)
ratio_w = resize_w / float(w)
return img, [ratio_h, ratio_w]
class BaseRecLabelDecode(object):
""" Convert between text-label and text-index """
def __init__(self, config):
support_character_type = [
'ch', 'en', 'EN_symbol', 'french', 'german', 'japan', 'korean',
'it', 'xi', 'pu', 'ru', 'ar', 'ta', 'ug', 'fa', 'ur', 'rs', 'oc',
'rsc', 'bg', 'uk', 'be', 'te', 'ka', 'chinese_cht', 'hi', 'mr',
'ne', 'EN'
]
character_type = config['character_type']
character_dict_path = config['character_dict_path']
use_space_char = True
assert character_type in support_character_type, "Only {} are supported now but get {}".format(
support_character_type, character_type)
self.beg_str = "sos"
self.end_str = "eos"
if character_type == "en":
self.character_str = "0123456789abcdefghijklmnopqrstuvwxyz"
dict_character = list(self.character_str)
elif character_type == "EN_symbol":
# same with ASTER setting (use 94 char).
self.character_str = string.printable[:-6]
dict_character = list(self.character_str)
elif character_type in support_character_type:
self.character_str = ""
assert character_dict_path is not None, "character_dict_path should not be None when character_type is {}".format(
character_type)
with open(character_dict_path, "rb") as fin:
lines = fin.readlines()
for line in lines:
line = line.decode('utf-8').strip("\n").strip("\r\n")
self.character_str += line
if use_space_char:
self.character_str += " "
dict_character = list(self.character_str)
else:
raise NotImplementedError
self.character_type = character_type
dict_character = self.add_special_char(dict_character)
self.dict = {}
for i, char in enumerate(dict_character):
self.dict[char] = i
self.character = dict_character
def add_special_char(self, dict_character):
return dict_character
def decode(self, text_index, text_prob=None, is_remove_duplicate=False):
""" convert text-index into text-label. """
result_list = []
ignored_tokens = self.get_ignored_tokens()
batch_size = len(text_index)
for batch_idx in range(batch_size):
char_list = []
conf_list = []
for idx in range(len(text_index[batch_idx])):
if text_index[batch_idx][idx] in ignored_tokens:
continue
if is_remove_duplicate:
# only for predict
if idx > 0 and text_index[batch_idx][idx - 1] == text_index[
batch_idx][idx]:
continue
char_list.append(self.character[int(text_index[batch_idx][
idx])])
if text_prob is not None:
conf_list.append(text_prob[batch_idx][idx])
else:
conf_list.append(1)
text = ''.join(char_list)
result_list.append((text, np.mean(conf_list)))
return result_list
def get_ignored_tokens(self):
return [0] # for ctc blank
class CTCLabelDecode(BaseRecLabelDecode):
""" Convert between text-label and text-index """
def __init__(
self,
config,
#character_dict_path=None,
#character_type='ch',
#use_space_char=False,
**kwargs):
super(CTCLabelDecode, self).__init__(config)
def __call__(self, preds, label=None, *args, **kwargs):
preds_idx = preds.argmax(axis=2)
preds_prob = preds.max(axis=2)
text = self.decode(preds_idx, preds_prob, is_remove_duplicate=True)
if label is None:
return text
label = self.decode(label)
return text, label
def add_special_char(self, dict_character):
dict_character = ['blank'] + dict_character
return dict_character
class CharacterOps(object):
""" Convert between text-label and text-index """
def __init__(self, config):
self.character_type = config['character_type']
self.loss_type = config['loss_type']
if self.character_type == "en":
self.character_str = "0123456789abcdefghijklmnopqrstuvwxyz"
dict_character = list(self.character_str)
elif self.character_type == "ch":
character_dict_path = config['character_dict_path']
self.character_str = ""
with open(character_dict_path, "rb") as fin:
lines = fin.readlines()
for line in lines:
line = line.decode('utf-8').strip("\n").strip("\r\n")
self.character_str += line
dict_character = list(self.character_str)
elif self.character_type == "en_sensitive":
# same with ASTER setting (use 94 char).
self.character_str = string.printable[:-6]
dict_character = list(self.character_str)
else:
self.character_str = None
assert self.character_str is not None, \
"Nonsupport type of the character: {}".format(self.character_str)
self.beg_str = "sos"
self.end_str = "eos"
if self.loss_type == "attention":
dict_character = [self.beg_str, self.end_str] + dict_character
self.dict = {}
for i, char in enumerate(dict_character):
self.dict[char] = i
self.character = dict_character
def encode(self, text):
"""convert text-label into text-index.
input:
text: text labels of each image. [batch_size]
output:
text: concatenated text index for CTCLoss.
[sum(text_lengths)] = [text_index_0 + text_index_1 + ... + text_index_(n - 1)]
length: length of each text. [batch_size]
"""
if self.character_type == "en":
text = text.lower()
text_list = []
for char in text:
if char not in self.dict:
continue
text_list.append(self.dict[char])
text = np.array(text_list)
return text
def decode(self, text_index, is_remove_duplicate=False):
""" convert text-index into text-label. """
char_list = []
char_num = self.get_char_num()
if self.loss_type == "attention":
beg_idx = self.get_beg_end_flag_idx("beg")
end_idx = self.get_beg_end_flag_idx("end")
ignored_tokens = [beg_idx, end_idx]
else:
ignored_tokens = [char_num]
for idx in range(len(text_index)):
if text_index[idx] in ignored_tokens:
continue
if is_remove_duplicate:
if idx > 0 and text_index[idx - 1] == text_index[idx]:
continue
char_list.append(self.character[text_index[idx]])
text = ''.join(char_list)
return text
def get_char_num(self):
return len(self.character)
def get_beg_end_flag_idx(self, beg_or_end):
if self.loss_type == "attention":
if beg_or_end == "beg":
idx = np.array(self.dict[self.beg_str])
elif beg_or_end == "end":
idx = np.array(self.dict[self.end_str])
else:
assert False, "Unsupport type %s in get_beg_end_flag_idx"\
% beg_or_end
return idx
else:
err = "error in get_beg_end_flag_idx when using the loss %s"\
% (self.loss_type)
assert False, err
class OCRReader(object):
def __init__(self,
algorithm="CRNN",
image_shape=[3, 32, 320],
char_type="ch",
batch_num=1,
char_dict_path="./ppocr_keys_v1.txt"):
self.rec_image_shape = image_shape
self.character_type = char_type
self.rec_batch_num = batch_num
char_ops_params = {}
char_ops_params["character_type"] = char_type
char_ops_params["character_dict_path"] = char_dict_path
char_ops_params['loss_type'] = 'ctc'
self.char_ops = CharacterOps(char_ops_params)
self.label_ops = CTCLabelDecode(char_ops_params)
def resize_norm_img(self, img, max_wh_ratio):
imgC, imgH, imgW = self.rec_image_shape
if self.character_type == "ch":
imgW = int(32 * max_wh_ratio)
h = img.shape[0]
w = img.shape[1]
ratio = w / float(h)
if math.ceil(imgH * ratio) > imgW:
resized_w = imgW
else:
resized_w = int(math.ceil(imgH * ratio))
resized_image = cv2.resize(img, (resized_w, imgH))
resized_image = resized_image.astype('float32')
resized_image = resized_image.transpose((2, 0, 1)) / 255
resized_image -= 0.5
resized_image /= 0.5
padding_im = np.zeros((imgC, imgH, imgW), dtype=np.float32)
padding_im[:, :, 0:resized_w] = resized_image
return padding_im
def preprocess(self, img_list):
img_num = len(img_list)
norm_img_batch = []
max_wh_ratio = 0
for ino in range(img_num):
h, w = img_list[ino].shape[0:2]
wh_ratio = w * 1.0 / h
max_wh_ratio = max(max_wh_ratio, wh_ratio)
for ino in range(img_num):
norm_img = self.resize_norm_img(img_list[ino], max_wh_ratio)
norm_img = norm_img[np.newaxis, :]
norm_img_batch.append(norm_img)
norm_img_batch = np.concatenate(norm_img_batch)
norm_img_batch = norm_img_batch.copy()
return norm_img_batch[0]
def postprocess_old(self, outputs, with_score=False):
rec_res = []
rec_idx_lod = outputs["ctc_greedy_decoder_0.tmp_0.lod"]
rec_idx_batch = outputs["ctc_greedy_decoder_0.tmp_0"]
if with_score:
predict_lod = outputs["softmax_0.tmp_0.lod"]
for rno in range(len(rec_idx_lod) - 1):
beg = rec_idx_lod[rno]
end = rec_idx_lod[rno + 1]
if isinstance(rec_idx_batch, list):
rec_idx_tmp = [x[0] for x in rec_idx_batch[beg:end]]
else: #nd array
rec_idx_tmp = rec_idx_batch[beg:end, 0]
preds_text = self.char_ops.decode(rec_idx_tmp)
if with_score:
beg = predict_lod[rno]
end = predict_lod[rno + 1]
if isinstance(outputs["softmax_0.tmp_0"], list):
outputs["softmax_0.tmp_0"] = np.array(outputs[
"softmax_0.tmp_0"]).astype(np.float32)
probs = outputs["softmax_0.tmp_0"][beg:end, :]
ind = np.argmax(probs, axis=1)
blank = probs.shape[1]
valid_ind = np.where(ind != (blank - 1))[0]
score = np.mean(probs[valid_ind, ind[valid_ind]])
rec_res.append([preds_text, score])
else:
rec_res.append([preds_text])
return rec_res
def postprocess(self, outputs, with_score=False):
preds = outputs["save_infer_model/scale_0.tmp_1"]
try:
preds = preds.numpy()
except:
pass
preds_idx = preds.argmax(axis=2)
preds_prob = preds.max(axis=2)
text = self.label_ops.decode(
preds_idx, preds_prob, is_remove_duplicate=True)
return text
deploy/pdserving/win/ocr_web_client.py 0 → 100644
View file @ 4684103f
# Copyright (c) 2020 PaddlePaddle 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.
# -*- coding: utf-8 -*-
import requests
import json
import cv2
import base64
import os, sys
import time
def cv2_to_base64(image):
#data = cv2.imencode('.jpg', image)[1]
return base64.b64encode(image).decode(
'utf8') #data.tostring()).decode('utf8')
headers = {"Content-type": "application/json"}
url = "http://127.0.0.1:9292/ocr/prediction"
test_img_dir = "../../../doc/imgs/"
for idx, img_file in enumerate(os.listdir(test_img_dir)):
with open(os.path.join(test_img_dir, img_file), 'rb') as file:
image_data1 = file.read()
image = cv2_to_base64(image_data1)
for i in range(1):
data = {"feed": [{"image": image}], "fetch": ["save_infer_model/scale_0.tmp_1"]}
r = requests.post(url=url, headers=headers, data=json.dumps(data))
print(r.json())
test_img_dir = "../../../doc/imgs/"
print("==> total number of test imgs: ", len(os.listdir(test_img_dir)))
deploy/pdserving/win/ocr_web_server.py 0 → 100644
View file @ 4684103f
# Copyright (c) 2020 PaddlePaddle 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.
from paddle_serving_client import Client
import cv2
import sys
import numpy as np
import os
from paddle_serving_client import Client
from paddle_serving_app.reader import Sequential, URL2Image, ResizeByFactor
from paddle_serving_app.reader import Div, Normalize, Transpose
from paddle_serving_app.reader import DBPostProcess, FilterBoxes, GetRotateCropImage, SortedBoxes
from ocr_reader import OCRReader
try:
from paddle_serving_server_gpu.web_service import WebService
except ImportError:
from paddle_serving_server.web_service import WebService
from paddle_serving_app.local_predict import LocalPredictor
import time
import re
import base64
class OCRService(WebService):
def init_det_debugger(self, det_model_config):
self.det_preprocess = Sequential([
ResizeByFactor(32, 960), Div(255),
Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]), Transpose(
(2, 0, 1))
])
self.det_client = LocalPredictor()
if sys.argv[1] == 'gpu':
self.det_client.load_model_config(
det_model_config, use_gpu=True, gpu_id=0)
elif sys.argv[1] == 'cpu':
self.det_client.load_model_config(det_model_config)
self.ocr_reader = OCRReader(
char_dict_path="../../../ppocr/utils/ppocr_keys_v1.txt")
def preprocess(self, feed=[], fetch=[]):
data = base64.b64decode(feed[0]["image"].encode('utf8'))
data = np.fromstring(data, np.uint8)
im = cv2.imdecode(data, cv2.IMREAD_COLOR)
ori_h, ori_w, _ = im.shape
det_img = self.det_preprocess(im)
_, new_h, new_w = det_img.shape
det_img = det_img[np.newaxis, :]
det_img = det_img.copy()
det_out = self.det_client.predict(
feed={"x": det_img}, fetch=["save_infer_model/scale_0.tmp_1"], batch=True)
filter_func = FilterBoxes(10, 10)
post_func = DBPostProcess({
"thresh": 0.3,
"box_thresh": 0.5,
"max_candidates": 1000,
"unclip_ratio": 1.5,
"min_size": 3
})
sorted_boxes = SortedBoxes()
ratio_list = [float(new_h) / ori_h, float(new_w) / ori_w]
dt_boxes_list = post_func(det_out["save_infer_model/scale_0.tmp_1"], [ratio_list])
dt_boxes = filter_func(dt_boxes_list[0], [ori_h, ori_w])
dt_boxes = sorted_boxes(dt_boxes)
get_rotate_crop_image = GetRotateCropImage()
img_list = []
max_wh_ratio = 0
for i, dtbox in enumerate(dt_boxes):
boximg = get_rotate_crop_image(im, dt_boxes[i])
img_list.append(boximg)
h, w = boximg.shape[0:2]
wh_ratio = w * 1.0 / h
max_wh_ratio = max(max_wh_ratio, wh_ratio)
if len(img_list) == 0:
return [], []
_, w, h = self.ocr_reader.resize_norm_img(img_list[0],
max_wh_ratio).shape
imgs = np.zeros((len(img_list), 3, w, h)).astype('float32')
for id, img in enumerate(img_list):
norm_img = self.ocr_reader.resize_norm_img(img, max_wh_ratio)
imgs[id] = norm_img
feed = {"x": imgs.copy()}
fetch = ["save_infer_model/scale_0.tmp_1"]
return feed, fetch, True
def postprocess(self, feed={}, fetch=[], fetch_map=None):
rec_res = self.ocr_reader.postprocess(fetch_map, with_score=True)
res_lst = []
for res in rec_res:
res_lst.append(res[0])
res = {"res": res_lst}
return res
ocr_service = OCRService(name="ocr")
ocr_service.load_model_config("../ppocr_rec_mobile_2.0_serving")
ocr_service.prepare_server(workdir="workdir", port=9292)
ocr_service.init_det_debugger(det_model_config="../ppocr_det_mobile_2.0_serving")
if sys.argv[1] == 'gpu':
ocr_service.set_gpus("0")
ocr_service.run_debugger_service(gpu=True)
elif sys.argv[1] == 'cpu':
ocr_service.run_debugger_service()
ocr_service.run_web_service()
deploy/slim/prune/README.md
View file @ 4684103f
...@@ -23,13 +23,13 @@ ...@@ -23,13 +23,13 @@
```bash ```bash
git clone https://github.com/PaddlePaddle/PaddleSlim.git git clone https://github.com/PaddlePaddle/PaddleSlim.git
cd PaddleSlim
git checkout develop git checkout develop
cd Paddleslim
python3 setup.py install python3 setup.py install
``` ```
### 2. 获取预训练模型 ### 2. 获取预训练模型
模型裁剪需要加载事先训练好的模型，PaddleOCR也提供了一系列(模型)[../../../doc/doc_ch/models_list.md]，开发者可根据需要自行选择模型或使用自己的模型。 模型裁剪需要加载事先训练好的模型，PaddleOCR也提供了一系列[模型](../../../doc/doc_ch/models_list.md)，开发者可根据需要自行选择模型或使用自己的模型。
### 3. 敏感度分析训练 ### 3. 敏感度分析训练
...@@ -49,14 +49,14 @@ python3 setup.py install ...@@ -49,14 +49,14 @@ python3 setup.py install
进入PaddleOCR根目录，通过以下命令对模型进行敏感度分析训练： 进入PaddleOCR根目录，通过以下命令对模型进行敏感度分析训练：
```bash ```bash
python3.7 deploy/slim/prune/sensitivity_anal.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrain_weights="your trained model" python3.7 deploy/slim/prune/sensitivity_anal.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model="your trained model" Global.save_model_dir=./output/prune_model/
``` ```
### 4. 导出模型、预测部署 ### 4. 导出模型、预测部署
在得到裁剪训练保存的模型后，我们可以将其导出为inference_model： 在得到裁剪训练保存的模型后，我们可以将其导出为inference_model：
```bash ```bash
pytho3.7 deploy/slim/prune/export_prune_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrain_weights=./output/det_db/best_accuracy Global.save_inference_dir=inference_model pytho3.7 deploy/slim/prune/export_prune_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./output/det_db/best_accuracy Global.save_inference_dir=./prune/prune_inference_model
``` ```
inference model的预测和部署参考： inference model的预测和部署参考：
......
deploy/slim/prune/README_en.md
View file @ 4684103f
...@@ -22,15 +22,15 @@ Five steps for OCR model prune: ...@@ -22,15 +22,15 @@ Five steps for OCR model prune:
```bash ```bash
git clone https://github.com/PaddlePaddle/PaddleSlim.git git clone https://github.com/PaddlePaddle/PaddleSlim.git
cd PaddleSlim
git checkout develop git checkout develop
cd Paddleslim
python3 setup.py install python3 setup.py install
``` ```
### 2. Download Pretrain Model ### 2. Download Pretrain Model
Model prune needs to load pre-trained models. Model prune needs to load pre-trained models.
PaddleOCR also provides a series of (models)[../../../doc/doc_en/models_list_en.md]. Developers can choose their own models or use their own models according to their needs. PaddleOCR also provides a series of [models](../../../doc/doc_en/models_list_en.md). Developers can choose their own models or use their own models according to their needs.
### 3. Pruning sensitivity analysis ### 3. Pruning sensitivity analysis
...@@ -54,7 +54,7 @@ Enter the PaddleOCR root directory，perform sensitivity analysis on the model w ...@@ -54,7 +54,7 @@ Enter the PaddleOCR root directory，perform sensitivity analysis on the model w
```bash ```bash
python3.7 deploy/slim/prune/sensitivity_anal.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrain_weights="your trained model" python3.7 deploy/slim/prune/sensitivity_anal.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model="your trained model" Global.save_model_dir=./output/prune_model/
``` ```
...@@ -63,7 +63,7 @@ python3.7 deploy/slim/prune/sensitivity_anal.py -c configs/det/ch_ppocr_v2.0/ch_ ...@@ -63,7 +63,7 @@ python3.7 deploy/slim/prune/sensitivity_anal.py -c configs/det/ch_ppocr_v2.0/ch_
We can export the pruned model as inference_model for deployment: We can export the pruned model as inference_model for deployment:
```bash ```bash
python deploy/slim/prune/export_prune_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrain_weights=./output/det_db/best_accuracy Global.test_batch_size_per_card=1 Global.save_inference_dir=inference_model python deploy/slim/prune/export_prune_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./output/det_db/best_accuracy Global.save_inference_dir=./prune/prune_inference_model
``` ```
Reference for prediction and deployment of inference model: Reference for prediction and deployment of inference model:
......
deploy/slim/quantization/README.md
View file @ 4684103f
...@@ -23,7 +23,7 @@ ...@@ -23,7 +23,7 @@
```bash ```bash
git clone https://github.com/PaddlePaddle/PaddleSlim.git git clone https://github.com/PaddlePaddle/PaddleSlim.git
cd Paddleslim cd PaddleSlim
python setup.py install python setup.py install
``` ```
...@@ -37,12 +37,12 @@ PaddleOCR提供了一系列训练好的[模型](../../../doc/doc_ch/models_list. ...@@ -37,12 +37,12 @@ PaddleOCR提供了一系列训练好的[模型](../../../doc/doc_ch/models_list.
量化训练的代码位于slim/quantization/quant.py 中，比如训练检测模型，训练指令如下： 量化训练的代码位于slim/quantization/quant.py 中，比如训练检测模型，训练指令如下：
```bash ```bash
python deploy/slim/quantization/quant.py -c configs/det/det_mv3_db.yml -o Global.pretrain_weights='your trained model' Global.save_model_dir=./output/quant_model python deploy/slim/quantization/quant.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model='your trained model' Global.save_model_dir=./output/quant_model
# 比如下载提供的训练模型 # 比如下载提供的训练模型
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar
tar -xf ch_ppocr_mobile_v2.0_det_train.tar tar -xf ch_ppocr_mobile_v2.0_det_train.tar
python deploy/slim/quantization/quant.py -c configs/det/det_mv3_db.yml -o Global.pretrain_weights=./ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.save_inference_dir=./output/quant_inference_model python deploy/slim/quantization/quant.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.save_model_dir=./output/quant_inference_model
``` ```
如果要训练识别模型的量化，修改配置文件和加载的模型参数即可。 如果要训练识别模型的量化，修改配置文件和加载的模型参数即可。
...@@ -52,7 +52,7 @@ python deploy/slim/quantization/quant.py -c configs/det/det_mv3_db.yml -o Global ...@@ -52,7 +52,7 @@ python deploy/slim/quantization/quant.py -c configs/det/det_mv3_db.yml -o Global
在得到量化训练保存的模型后，我们可以将其导出为inference_model，用于预测部署： 在得到量化训练保存的模型后，我们可以将其导出为inference_model，用于预测部署：
```bash ```bash
python deploy/slim/quantization/export_model.py -c configs/det/det_mv3_db.yml -o Global.checkpoints=output/quant_model/best_accuracy Global.save_model_dir=./output/quant_inference_model python deploy/slim/quantization/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.checkpoints=output/quant_model/best_accuracy Global.save_inference_dir=./output/quant_inference_model
``` ```
### 5. 量化模型部署 ### 5. 量化模型部署
......
deploy/slim/quantization/README_en.md
View file @ 4684103f
...@@ -26,7 +26,7 @@ After training, if you want to further compress the model size and accelerate th ...@@ -26,7 +26,7 @@ After training, if you want to further compress the model size and accelerate th
```bash ```bash
git clone https://github.com/PaddlePaddle/PaddleSlim.git git clone https://github.com/PaddlePaddle/PaddleSlim.git
cd Paddleslim cd PaddlSlim
python setup.py install python setup.py install
``` ```
...@@ -43,13 +43,12 @@ After the quantization strategy is defined, the model can be quantified. ...@@ -43,13 +43,12 @@ After the quantization strategy is defined, the model can be quantified.
The code for quantization training is located in `slim/quantization/quant.py`. For example, to train a detection model, the training instructions are as follows: The code for quantization training is located in `slim/quantization/quant.py`. For example, to train a detection model, the training instructions are as follows:
```bash ```bash
python deploy/slim/quantization/quant.py -c configs/det/det_mv3_db.yml -o Global.pretrain_weights='your trained model' Global.save_model_dir=./output/quant_model python deploy/slim/quantization/quant.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model='your trained model' Global.save_model_dir=./output/quant_model
# download provided model # download provided model
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar
tar -xf ch_ppocr_mobile_v2.0_det_train.tar tar -xf ch_ppocr_mobile_v2.0_det_train.tar
python deploy/slim/quantization/quant.py -c configs/det/det_mv3_db.yml -o Global.pretrain_weights=./ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.save_model_dir=./output/quant_model python deploy/slim/quantization/quant.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.save_model_dir=./output/quant_model
``` ```
...@@ -58,7 +57,7 @@ python deploy/slim/quantization/quant.py -c configs/det/det_mv3_db.yml -o Global ...@@ -58,7 +57,7 @@ python deploy/slim/quantization/quant.py -c configs/det/det_mv3_db.yml -o Global
After getting the model after pruning and finetuning we, can export it as inference_model for predictive deployment: After getting the model after pruning and finetuning we, can export it as inference_model for predictive deployment:
```bash ```bash
python deploy/slim/quantization/export_model.py -c configs/det/det_mv3_db.yml -o Global.checkpoints=output/quant_model/best_accuracy Global.save_inference_dir=./output/quant_inference_model python deploy/slim/quantization/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.checkpoints=output/quant_model/best_accuracy Global.save_inference_dir=./output/quant_inference_model
``` ```
### 5. Deploy ### 5. Deploy
......
deploy/slim/quantization/quant.py
View file @ 4684103f
...@@ -112,10 +112,6 @@ def main(config, device, logger, vdl_writer): ...@@ -112,10 +112,6 @@ def main(config, device, logger, vdl_writer):
config['Architecture']["Head"]['out_channels'] = char_num config['Architecture']["Head"]['out_channels'] = char_num
model = build_model(config['Architecture']) model = build_model(config['Architecture'])
# prepare to quant
quanter = QAT(config=quant_config, act_preprocess=PACT)
quanter.quantize(model)
if config['Global']['distributed']: if config['Global']['distributed']:
model = paddle.DataParallel(model) model = paddle.DataParallel(model)
...@@ -136,31 +132,15 @@ def main(config, device, logger, vdl_writer): ...@@ -136,31 +132,15 @@ def main(config, device, logger, vdl_writer):
logger.info('train dataloader has {} iters, valid dataloader has {} iters'. logger.info('train dataloader has {} iters, valid dataloader has {} iters'.
format(len(train_dataloader), len(valid_dataloader))) format(len(train_dataloader), len(valid_dataloader)))
quanter = QAT(config=quant_config, act_preprocess=PACT)
quanter.quantize(model)
# start train # start train
program.train(config, train_dataloader, valid_dataloader, device, model, program.train(config, train_dataloader, valid_dataloader, device, model,
loss_class, optimizer, lr_scheduler, post_process_class, loss_class, optimizer, lr_scheduler, post_process_class,
eval_class, pre_best_model_dict, logger, vdl_writer) eval_class, pre_best_model_dict, logger, vdl_writer)
def test_reader(config, device, logger):
loader = build_dataloader(config, 'Train', device, logger)
import time
starttime = time.time()
count = 0
try:
for data in loader():
count += 1
if count % 1 == 0:
batch_time = time.time() - starttime
starttime = time.time()
logger.info("reader: {}, {}, {}".format(
count, len(data[0]), batch_time))
except Exception as e:
logger.info(e)
logger.info("finish reader: {}, Success!".format(count))
if __name__ == '__main__': if __name__ == '__main__':
config, device, logger, vdl_writer = program.preprocess(is_train=True) config, device, logger, vdl_writer = program.preprocess(is_train=True)
main(config, device, logger, vdl_writer) main(config, device, logger, vdl_writer)
# test_reader(config, device, logger)
doc/doc_ch/detection.md
View file @ 4684103f
...@@ -45,26 +45,17 @@ json.dumps编码前的图像标注信息是包含多个字典的list，字典中 ...@@ -45,26 +45,17 @@ json.dumps编码前的图像标注信息是包含多个字典的list，字典中
## 快速启动训练 ## 快速启动训练
首先下载模型backbone的pretrain model，PaddleOCR的检测模型目前支持两种backbone，分别是MobileNetV3、ResNet_vd系列， 首先下载模型backbone的pretrain model，PaddleOCR的检测模型目前支持两种backbone，分别是MobileNetV3、ResNet_vd系列，
您可以根据需求使用[PaddleClas](https://github.com/PaddlePaddle/PaddleClas/tree/master/ppcls/modeling/architectures)中的模型更换backbone。 您可以根据需求使用[PaddleClas](https://github.com/PaddlePaddle/PaddleClas/tree/develop/ppcls/modeling/architectures)中的模型更换backbone，
对应的backbone预训练模型可以从[PaddleClas repo 主页中找到下载链接](https://github.com/PaddlePaddle/PaddleClas#mobile-series)。
```shell ```shell
cd PaddleOCR/ cd PaddleOCR/
# 根据backbone的不同选择下载对应的预训练模型
# 下载MobileNetV3的预训练模型 # 下载MobileNetV3的预训练模型
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_large_x0_5_pretrained.tar wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/MobileNetV3_large_x0_5_pretrained.pdparams
# 或，下载ResNet18_vd的预训练模型 # 或，下载ResNet18_vd的预训练模型
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/ResNet18_vd_pretrained.tar wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ResNet18_vd_pretrained.pdparams
# 或，下载ResNet50_vd的预训练模型 # 或，下载ResNet50_vd的预训练模型
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_ssld_pretrained.tar wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ResNet50_vd_ssld_pretrained.pdparams
# 解压预训练模型文件，以MobileNetV3为例
tar -xf ./pretrain_models/MobileNetV3_large_x0_5_pretrained.tar ./pretrain_models/
# 注：正确解压backbone预训练权重文件后，文件夹下包含众多以网络层命名的权重文件，格式如下：
./pretrain_models/MobileNetV3_large_x0_5_pretrained/
└─ conv_last_bn_mean
└─ conv_last_bn_offset
└─ conv_last_bn_scale
└─ conv_last_bn_variance
└─ ......
``` ```
...@@ -120,16 +111,16 @@ python3 tools/eval.py -c configs/det/det_mv3_db.yml -o Global.checkpoints="{pat ...@@ -120,16 +111,16 @@ python3 tools/eval.py -c configs/det/det_mv3_db.yml -o Global.checkpoints="{pat
测试单张图像的检测效果 测试单张图像的检测效果
```shell ```shell
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy"
``` ```
测试DB模型时，调整后处理阈值， 测试DB模型时，调整后处理阈值，
```shell ```shell
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5 python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5
``` ```
测试文件夹下所有图像的检测效果 测试文件夹下所有图像的检测效果
```shell ```shell
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/det_db/best_accuracy"
``` ```
doc/doc_ch/distributed_training.md 0 → 100644
View file @ 4684103f
# 分布式训练
## 简介
* 分布式训练的高性能，是飞桨的核心优势技术之一，在分类任务上，分布式训练可以达到几乎线性的加速比。OCR训练任务中往往包含大量训练数据，以识别为例，ppocrv2.0模型在训练时使用了1800W数据，如果使用单机训练，会非常耗时。因此，PaddleOCR中使用分布式训练接口完成训练任务，同时支持单机训练与多机训练。更多关于分布式训练的方法与文档可以参考：[分布式训练快速开始教程](https://fleet-x.readthedocs.io/en/latest/paddle_fleet_rst/parameter_server/ps_quick_start.html)。
## 使用方法
### 单机训练
* 以识别为例，本地准备好数据之后，使用`paddle.distributed.launch`的接口启动训练任务即可。下面为运行代码示例。
```shell
python3 -m paddle.distributed.launch \
--log_dir=./log/ \
--gpus '0,1,2,3,4,5,6,7' \
tools/train.py \
-c configs/rec/rec_mv3_none_bilstm_ctc.yml
```
### 多机训练
* 相比单机训练，多机训练时，只需要添加`--ips`的参数，该参数表示需要参与分布式训练的机器的ip列表，不同机器的ip用逗号隔开。下面为运行代码示例。
```shell
ip_list="192.168.0.1,192.168.0.2"
python3 -m paddle.distributed.launch \
--log_dir=./log/ \
--ips="${ip_list}" \
--gpus="0,1,2,3,4,5,6,7" \
tools/train.py \
-c configs/rec/rec_mv3_none_bilstm_ctc.yml
```
**注：**
* 不同机器的ip信息需要用逗号隔开，可以通过`ifconfig`或者`ipconfig`查看。
* 不同机器之间需要做免密设置，且可以直接ping通，否则无法完成通信。
* 不同机器之间的代码、数据与运行命令或脚本需要保持一致，且所有的机器上都需要运行设置好的训练命令或者脚本。最终`ip_list`中的第一台机器的第一块设备是trainer0，以此类推。
## 性能效果测试
* 基于单机8卡P40，和2机8卡P40，在26W公开识别数据集(LSVT, RCTW, MTWI)上进行训练，最终耗时如下。
| 模型 | 配置文件 | 机器数量 | 每台机器的GPU数量 | 训练时间 | 识别Acc | 加速比 |
| :----------------------: | :------------: | :------------: | :---------------: | :----------: | :-----------: | :-----------: |
| CRNN | configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml | 1 | 8 | 60h | 66.7% | - |
| CRNN | configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml | 2 | 8 | 40h | 67.0% | 150% |
可以看出，精度没有下降的情况下，训练时间由60h缩短为了40h，加速比可以达到60h/40h=150%，效率为60h/(40h*2)=75%。
doc/doc_ch/inference.md
View file @ 4684103f
...@@ -49,10 +49,9 @@ wget -P ./ch_lite/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobi ...@@ -49,10 +49,9 @@ wget -P ./ch_lite/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobi
# -c 后面设置训练算法的yml配置文件 # -c 后面设置训练算法的yml配置文件
# -o 配置可选参数 # -o 配置可选参数
# Global.pretrained_model 参数设置待转换的训练模型地址，不用添加文件后缀 .pdmodel，.pdopt或.pdparams。 # Global.pretrained_model 参数设置待转换的训练模型地址，不用添加文件后缀 .pdmodel，.pdopt或.pdparams。
# Global.load_static_weights 参数需要设置为 False。
# Global.save_inference_dir参数设置转换的模型将保存的地址。 # Global.save_inference_dir参数设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_db/ python3 tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.save_inference_dir=./inference/det_db/
``` ```
转inference模型时，使用的配置文件和训练时使用的配置文件相同。另外，还需要设置配置文件中的`Global.pretrained_model`参数，其指向训练中保存的模型参数文件。 转inference模型时，使用的配置文件和训练时使用的配置文件相同。另外，还需要设置配置文件中的`Global.pretrained_model`参数，其指向训练中保存的模型参数文件。
转换成功后，在模型保存目录下有三个文件： 转换成功后，在模型保存目录下有三个文件：
...@@ -76,10 +75,9 @@ wget -P ./ch_lite/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobi ...@@ -76,10 +75,9 @@ wget -P ./ch_lite/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobi
# -c 后面设置训练算法的yml配置文件 # -c 后面设置训练算法的yml配置文件
# -o 配置可选参数 # -o 配置可选参数
# Global.pretrained_model 参数设置待转换的训练模型地址，不用添加文件后缀 .pdmodel，.pdopt或.pdparams。 # Global.pretrained_model 参数设置待转换的训练模型地址，不用添加文件后缀 .pdmodel，.pdopt或.pdparams。
# Global.load_static_weights 参数需要设置为 False。
# Global.save_inference_dir参数设置转换的模型将保存的地址。 # Global.save_inference_dir参数设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_rec_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/rec_crnn/ python3 tools/export_model.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_rec_train/best_accuracy Global.save_inference_dir=./inference/rec_crnn/
``` ```
**注意：**如果您是在自己的数据集上训练的模型，并且调整了中文字符的字典文件，请注意修改配置文件中的`character_dict_path`是否是所需要的字典文件。 **注意：**如果您是在自己的数据集上训练的模型，并且调整了中文字符的字典文件，请注意修改配置文件中的`character_dict_path`是否是所需要的字典文件。
...@@ -105,10 +103,9 @@ wget -P ./ch_lite/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobi ...@@ -105,10 +103,9 @@ wget -P ./ch_lite/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobi
# -c 后面设置训练算法的yml配置文件 # -c 后面设置训练算法的yml配置文件
# -o 配置可选参数 # -o 配置可选参数
# Global.pretrained_model 参数设置待转换的训练模型地址，不用添加文件后缀 .pdmodel，.pdopt或.pdparams。 # Global.pretrained_model 参数设置待转换的训练模型地址，不用添加文件后缀 .pdmodel，.pdopt或.pdparams。
# Global.load_static_weights 参数需要设置为 False。
# Global.save_inference_dir参数设置转换的模型将保存的地址。 # Global.save_inference_dir参数设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_cls_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/cls/ python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_cls_train/best_accuracy Global.save_inference_dir=./inference/cls/
``` ```
转换成功后，在目录下有三个文件： 转换成功后，在目录下有三个文件：
...@@ -164,7 +161,7 @@ python3 tools/infer/predict_det.py --image_dir="./doc/imgs/2.jpg" --det_model_di ...@@ -164,7 +161,7 @@ python3 tools/infer/predict_det.py --image_dir="./doc/imgs/2.jpg" --det_model_di
首先将DB文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例（ [模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_db_v2.0_train.tar) )，可以使用如下命令进行转换： 首先将DB文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例（ [模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_db_v2.0_train.tar) )，可以使用如下命令进行转换：
``` ```
python3 tools/export_model.py -c configs/det/det_r50_vd_db.yml -o Global.pretrained_model=./det_r50_vd_db_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_db python3 tools/export_model.py -c configs/det/det_r50_vd_db.yml -o Global.pretrained_model=./det_r50_vd_db_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_db
``` ```
DB文本检测模型推理，可以执行如下命令： DB文本检测模型推理，可以执行如下命令：
...@@ -185,7 +182,7 @@ python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_ ...@@ -185,7 +182,7 @@ python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_
首先将EAST文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例（ [模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar) )，可以使用如下命令进行转换： 首先将EAST文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例（ [模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar) )，可以使用如下命令进行转换：
``` ```
python3 tools/export_model.py -c configs/det/det_r50_vd_east.yml -o Global.pretrained_model=./det_r50_vd_east_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_east python3 tools/export_model.py -c configs/det/det_r50_vd_east.yml -o Global.pretrained_model=./det_r50_vd_east_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_east
``` ```
**EAST文本检测模型推理，需要设置参数`--det_algorithm="EAST"`**，可以执行如下命令： **EAST文本检测模型推理，需要设置参数`--det_algorithm="EAST"`**，可以执行如下命令：
...@@ -205,7 +202,7 @@ python3 tools/infer/predict_det.py --det_algorithm="EAST" --image_dir="./doc/img ...@@ -205,7 +202,7 @@ python3 tools/infer/predict_det.py --det_algorithm="EAST" --image_dir="./doc/img
#### (1). 四边形文本检测模型（ICDAR2015） #### (1). 四边形文本检测模型（ICDAR2015）
首先将SAST文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例([模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_icdar15_v2.0_train.tar))，可以使用如下命令进行转换： 首先将SAST文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例([模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_icdar15_v2.0_train.tar))，可以使用如下命令进行转换：
``` ```
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_icdar15.yml -o Global.pretrained_model=./det_r50_vd_sast_icdar15_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_sast_ic15 python3 tools/export_model.py -c configs/det/det_r50_vd_sast_icdar15.yml -o Global.pretrained_model=./det_r50_vd_sast_icdar15_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_sast_ic15
``` ```
**SAST文本检测模型推理，需要设置参数`--det_algorithm="SAST"`**，可以执行如下命令： **SAST文本检测模型推理，需要设置参数`--det_algorithm="SAST"`**，可以执行如下命令：
...@@ -220,7 +217,7 @@ python3 tools/infer/predict_det.py --det_algorithm="SAST" --image_dir="./doc/img ...@@ -220,7 +217,7 @@ python3 tools/infer/predict_det.py --det_algorithm="SAST" --image_dir="./doc/img
首先将SAST文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在Total-Text英文数据集训练的模型为例（[模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_totaltext_v2.0_train.tar))，可以使用如下命令进行转换： 首先将SAST文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在Total-Text英文数据集训练的模型为例（[模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_totaltext_v2.0_train.tar))，可以使用如下命令进行转换：
``` ```
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_totaltext.yml -o Global.pretrained_model=./det_r50_vd_sast_totaltext_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_sast_tt python3 tools/export_model.py -c configs/det/det_r50_vd_sast_totaltext.yml -o Global.pretrained_model=./det_r50_vd_sast_totaltext_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_sast_tt
``` ```
...@@ -270,7 +267,7 @@ Predicts of ./doc/imgs_words/ch/word_4.jpg:('实力活力', 0.98458153) ...@@ -270,7 +267,7 @@ Predicts of ./doc/imgs_words/ch/word_4.jpg:('实力活力', 0.98458153)
的模型为例（ [模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/rec_r34_vd_none_bilstm_ctc_v2.0_train.tar) )，可以使用如下命令进行转换： 的模型为例（ [模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/rec_r34_vd_none_bilstm_ctc_v2.0_train.tar) )，可以使用如下命令进行转换：
``` ```
python3 tools/export_model.py -c configs/rec/rec_r34_vd_none_bilstm_ctc.yml -o Global.pretrained_model=./rec_r34_vd_none_bilstm_ctc_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/rec_crnn python3 tools/export_model.py -c configs/rec/rec_r34_vd_none_bilstm_ctc.yml -o Global.pretrained_model=./rec_r34_vd_none_bilstm_ctc_v2.0_train/best_accuracy Global.save_inference_dir=./inference/rec_crnn
``` ```
CRNN 文本识别模型推理，可以执行如下命令： CRNN 文本识别模型推理，可以执行如下命令：
......
doc/doc_ch/models_list.md
View file @ 4684103f
...@@ -31,7 +31,7 @@ PaddleOCR提供的可下载模型包括`推理模型`、`训练模型`、`预训 ...@@ -31,7 +31,7 @@ PaddleOCR提供的可下载模型包括`推理模型`、`训练模型`、`预训
|模型名称|模型简介|配置文件|推理模型大小|下载地址| |模型名称|模型简介|配置文件|推理模型大小|下载地址|
| --- | --- | --- | --- | --- | | --- | --- | --- | --- | --- |
|ch_ppocr_mobile_slim_v2.0_det|slim裁剪版超轻量模型，支持中英文、多语种文本检测|[ch_det_mv3_db_v2.0.yml](../../configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml)| |推理模型 (coming soon) / 训练模型 (coming soon)| |ch_ppocr_mobile_slim_v2.0_det|slim裁剪版超轻量模型，支持中英文、多语种文本检测|[ch_det_mv3_db_v2.0.yml](../../configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml)|2.6M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_det_prune_infer.tar)|
|ch_ppocr_mobile_v2.0_det|原始超轻量模型，支持中英文、多语种文本检测|[ch_det_mv3_db_v2.0.yml](../../configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml)|3M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar)| |ch_ppocr_mobile_v2.0_det|原始超轻量模型，支持中英文、多语种文本检测|[ch_det_mv3_db_v2.0.yml](../../configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml)|3M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar)|
|ch_ppocr_server_v2.0_det|通用模型，支持中英文、多语种文本检测，比超轻量模型更大，但效果更好|[ch_det_res18_db_v2.0.yml](../../configs/det/ch_ppocr_v2.0/ch_det_res18_db_v2.0.yml)|47M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_train.tar)| |ch_ppocr_server_v2.0_det|通用模型，支持中英文、多语种文本检测，比超轻量模型更大，但效果更好|[ch_det_res18_db_v2.0.yml](../../configs/det/ch_ppocr_v2.0/ch_det_res18_db_v2.0.yml)|47M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_train.tar)|
......
doc/doc_ch/multi_languages.md
View file @ 4684103f
...@@ -82,33 +82,37 @@ Paddleocr目前支持80个语种，可以通过修改--lang参数进行切换， ...@@ -82,33 +82,37 @@ Paddleocr目前支持80个语种，可以通过修改--lang参数进行切换，
``` bash ``` bash
paddleocr --image_dir doc/imgs/japan_2.jpg --lang=japan paddleocr --image_dir doc/imgs_en/254.jpg --lang=en
``` ```
<div align="center"> <div align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/PaddleOCR/release/2.1/doc/imgs/japan_2.jpg" width="800"> <img src="../imgs_en/254.jpg" width="300" height="600">
<img src="../imgs_results/multi_lang/img_02.jpg" width="600" height="600">
</div> </div>
结果是一个list，每个item包含了文本框，文字和识别置信度 结果是一个list，每个item包含了文本框，文字和识别置信度
```text ```text
[[[671.0, 60.0], [847.0, 63.0], [847.0, 104.0], [671.0, 102.0]], ('もちもち', 0.9993342)] [('PHO CAPITAL', 0.95723116), [[66.0, 50.0], [327.0, 44.0], [327.0, 76.0], [67.0, 82.0]]]
[[[394.0, 82.0], [536.0, 77.0], [538.0, 127.0], [396.0, 132.0]], ('天然の', 0.9919842)] [('107 State Street', 0.96311164), [[72.0, 90.0], [451.0, 84.0], [452.0, 116.0], [73.0, 121.0]]]
[[[880.0, 89.0], [1014.0, 93.0], [1013.0, 127.0], [879.0, 124.0]], ('とろっと', 0.9976762)] [('Montpelier Vermont', 0.97389287), [[69.0, 132.0], [501.0, 126.0], [501.0, 158.0], [70.0, 164.0]]]
[[[1067.0, 101.0], [1294.0, 101.0], [1294.0, 138.0], [1067.0, 138.0]], ('後味のよい', 0.9988712)] [('8022256183', 0.99810505), [[71.0, 175.0], [363.0, 170.0], [364.0, 202.0], [72.0, 207.0]]]
[('REG 07-24-201706:59 PM', 0.93537045), [[73.0, 299.0], [653.0, 281.0], [654.0, 318.0], [74.0, 336.0]]]
[('045555', 0.99346405), [[509.0, 331.0], [651.0, 325.0], [652.0, 356.0], [511.0, 362.0]]]
[('CT1', 0.9988654), [[535.0, 367.0], [654.0, 367.0], [654.0, 406.0], [535.0, 406.0]]]
...... ......
``` ```
* 识别预测 * 识别预测
```bash ```bash
paddleocr --image_dir doc/imgs_words/japan/1.jpg --det false --lang=japan paddleocr --image_dir doc/imgs_words_en/word_308.png --det false --lang=en
``` ```
结果是一个tuple，返回识别结果和识别置信度 结果是一个tuple，返回识别结果和识别置信度
```text ```text
('したがって', 0.99965394) (0.99879867, 'LITTLE')
``` ```
* 检测预测 * 检测预测
......
doc/doc_en/detection_en.md
View file @ 4684103f
...@@ -38,28 +38,17 @@ If you want to train PaddleOCR on other datasets, please build the annotation fi ...@@ -38,28 +38,17 @@ If you want to train PaddleOCR on other datasets, please build the annotation fi
## TRAINING ## TRAINING
First download the pretrained model. The detection model of PaddleOCR currently supports 3 backbones, namely MobileNetV3, ResNet18_vd and ResNet50_vd. You can use the model in [PaddleClas](https://github.com/PaddlePaddle/PaddleClas/tree/master/ppcls/modeling/architectures) to replace backbone according to your needs. First download the pretrained model. The detection model of PaddleOCR currently supports 3 backbones, namely MobileNetV3, ResNet18_vd and ResNet50_vd. You can use the model in [PaddleClas](https://github.com/PaddlePaddle/PaddleClas/tree/develop/ppcls/modeling/architectures) to replace backbone according to your needs.
And the responding download link of backbone pretrain weights can be found in [PaddleClas repo](https://github.com/PaddlePaddle/PaddleClas#mobile-series).
```shell ```shell
cd PaddleOCR/ cd PaddleOCR/
# Download the pre-trained model of MobileNetV3 # Download the pre-trained model of MobileNetV3
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_large_x0_5_pretrained.tar wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/MobileNetV3_large_x0_5_pretrained.pdparams
# or, download the pre-trained model of ResNet18_vd # or, download the pre-trained model of ResNet18_vd
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/ResNet18_vd_pretrained.tar wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ResNet18_vd_pretrained.pdparams
# or, download the pre-trained model of ResNet50_vd # or, download the pre-trained model of ResNet50_vd
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_ssld_pretrained.tar wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ResNet50_vd_ssld_pretrained.pdparams
# decompressing the pre-training model file, take MobileNetV3 as an example
tar -xf ./pretrain_models/MobileNetV3_large_x0_5_pretrained.tar ./pretrain_models/
# Note: After decompressing the backbone pre-training weight file correctly, the file list in the folder is as follows:
./pretrain_models/MobileNetV3_large_x0_5_pretrained/
└─ conv_last_bn_mean
└─ conv_last_bn_offset
└─ conv_last_bn_scale
└─ conv_last_bn_variance
└─ ......
```
#### START TRAINING #### START TRAINING
*If CPU version installed, please set the parameter `use_gpu` to `false` in the configuration.* *If CPU version installed, please set the parameter `use_gpu` to `false` in the configuration.*
...@@ -113,16 +102,16 @@ python3 tools/eval.py -c configs/det/det_mv3_db.yml -o Global.checkpoints="{pat ...@@ -113,16 +102,16 @@ python3 tools/eval.py -c configs/det/det_mv3_db.yml -o Global.checkpoints="{pat
Test the detection result on a single image: Test the detection result on a single image:
```shell ```shell
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy"
``` ```
When testing the DB model, adjust the post-processing threshold: When testing the DB model, adjust the post-processing threshold:
```shell ```shell
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5 python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5
``` ```
Test the detection result on all images in the folder: Test the detection result on all images in the folder:
```shell ```shell
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/det_db/best_accuracy"
``` ```
doc/doc_en/inference_en.md
View file @ 4684103f
...@@ -52,10 +52,9 @@ The above model is a DB algorithm trained with MobileNetV3 as the backbone. To c ...@@ -52,10 +52,9 @@ The above model is a DB algorithm trained with MobileNetV3 as the backbone. To c
# -c Set the training algorithm yml configuration file # -c Set the training algorithm yml configuration file
# -o Set optional parameters # -o Set optional parameters
# Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams. # Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams.
# Global.load_static_weights needs to be set to False
# Global.save_inference_dir Set the address where the converted model will be saved. # Global.save_inference_dir Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_db/ python3 tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.save_inference_dir=./inference/det_db/
``` ```
When converting to an inference model, the configuration file used is the same as the configuration file used during training. In addition, you also need to set the `Global.pretrained_model` parameter in the configuration file. When converting to an inference model, the configuration file used is the same as the configuration file used during training. In addition, you also need to set the `Global.pretrained_model` parameter in the configuration file.
...@@ -80,10 +79,9 @@ The recognition model is converted to the inference model in the same way as the ...@@ -80,10 +79,9 @@ The recognition model is converted to the inference model in the same way as the
# -c Set the training algorithm yml configuration file # -c Set the training algorithm yml configuration file
# -o Set optional parameters # -o Set optional parameters
# Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams. # Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams.
# Global.load_static_weights needs to be set to False
# Global.save_inference_dir Set the address where the converted model will be saved. # Global.save_inference_dir Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_rec_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/rec_crnn/ python3 tools/export_model.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_rec_train/best_accuracy Global.save_inference_dir=./inference/rec_crnn/
``` ```
If you have a model trained on your own dataset with a different dictionary file, please make sure that you modify the `character_dict_path` in the configuration file to your dictionary file path. If you have a model trained on your own dataset with a different dictionary file, please make sure that you modify the `character_dict_path` in the configuration file to your dictionary file path.
...@@ -109,10 +107,9 @@ The angle classification model is converted to the inference model in the same w ...@@ -109,10 +107,9 @@ The angle classification model is converted to the inference model in the same w
# -c Set the training algorithm yml configuration file # -c Set the training algorithm yml configuration file
# -o Set optional parameters # -o Set optional parameters
# Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams. # Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams.
# Global.load_static_weights needs to be set to False
# Global.save_inference_dir Set the address where the converted model will be saved. # Global.save_inference_dir Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_cls_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/cls/ python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_cls_train/best_accuracy Global.save_inference_dir=./inference/cls/
``` ```
After the conversion is successful, there are two files in the directory: After the conversion is successful, there are two files in the directory:
...@@ -171,7 +168,7 @@ python3 tools/infer/predict_det.py --image_dir="./doc/imgs/22.jpg" --det_model_d ...@@ -171,7 +168,7 @@ python3 tools/infer/predict_det.py --image_dir="./doc/imgs/22.jpg" --det_model_d
First, convert the model saved in the DB text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_db_v2.0_train.tar)), you can use the following command to convert: First, convert the model saved in the DB text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_db_v2.0_train.tar)), you can use the following command to convert:
``` ```
python3 tools/export_model.py -c configs/det/det_r50_vd_db.yml -o Global.pretrained_model=./det_r50_vd_db_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_db python3 tools/export_model.py -c configs/det/det_r50_vd_db.yml -o Global.pretrained_model=./det_r50_vd_db_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_db
``` ```
DB text detection model inference, you can execute the following command: DB text detection model inference, you can execute the following command:
...@@ -192,7 +189,7 @@ The visualized text detection results are saved to the `./inference_results` fol ...@@ -192,7 +189,7 @@ The visualized text detection results are saved to the `./inference_results` fol
First, convert the model saved in the EAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar)), you can use the following command to convert: First, convert the model saved in the EAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar)), you can use the following command to convert:
``` ```
python3 tools/export_model.py -c configs/det/det_r50_vd_east.yml -o Global.pretrained_model=./det_r50_vd_east_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_east python3 tools/export_model.py -c configs/det/det_r50_vd_east.yml -o Global.pretrained_model=./det_r50_vd_east_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_east
``` ```
**For EAST text detection model inference, you need to set the parameter ``--det_algorithm="EAST"``**, run the following command: **For EAST text detection model inference, you need to set the parameter ``--det_algorithm="EAST"``**, run the following command:
...@@ -213,7 +210,7 @@ The visualized text detection results are saved to the `./inference_results` fol ...@@ -213,7 +210,7 @@ The visualized text detection results are saved to the `./inference_results` fol
First, convert the model saved in the SAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_icdar15_v2.0_train.tar)), you can use the following command to convert: First, convert the model saved in the SAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_icdar15_v2.0_train.tar)), you can use the following command to convert:
``` ```
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_icdar15.yml -o Global.pretrained_model=./det_r50_vd_sast_icdar15_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_sast_ic15 python3 tools/export_model.py -c configs/det/det_r50_vd_sast_icdar15.yml -o Global.pretrained_model=./det_r50_vd_sast_icdar15_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_sast_ic15
``` ```
**For SAST quadrangle text detection model inference, you need to set the parameter `--det_algorithm="SAST"`**, run the following command: **For SAST quadrangle text detection model inference, you need to set the parameter `--det_algorithm="SAST"`**, run the following command:
...@@ -230,7 +227,7 @@ The visualized text detection results are saved to the `./inference_results` fol ...@@ -230,7 +227,7 @@ The visualized text detection results are saved to the `./inference_results` fol
First, convert the model saved in the SAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the Total-Text English dataset as an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_totaltext_v2.0_train.tar)), you can use the following command to convert: First, convert the model saved in the SAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the Total-Text English dataset as an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_totaltext_v2.0_train.tar)), you can use the following command to convert:
``` ```
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_totaltext.yml -o Global.pretrained_model=./det_r50_vd_sast_totaltext_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_sast_tt python3 tools/export_model.py -c configs/det/det_r50_vd_sast_totaltext.yml -o Global.pretrained_model=./det_r50_vd_sast_totaltext_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_sast_tt
``` ```
**For SAST curved text detection model inference, you need to set the parameter `--det_algorithm="SAST"` and `--det_sast_polygon=True`**, run the following command: **For SAST curved text detection model inference, you need to set the parameter `--det_algorithm="SAST"` and `--det_sast_polygon=True`**, run the following command:
...@@ -279,7 +276,7 @@ Taking CRNN as an example, we introduce the recognition model inference based on ...@@ -279,7 +276,7 @@ Taking CRNN as an example, we introduce the recognition model inference based on
First, convert the model saved in the CRNN text recognition training process into an inference model. Taking the model based on Resnet34_vd backbone network, using MJSynth and SynthText (two English text recognition synthetic datasets) for training, as an example ([model download address](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/rec_r34_vd_none_bilstm_ctc_v2.0_train.tar)). It can be converted as follow: First, convert the model saved in the CRNN text recognition training process into an inference model. Taking the model based on Resnet34_vd backbone network, using MJSynth and SynthText (two English text recognition synthetic datasets) for training, as an example ([model download address](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/rec_r34_vd_none_bilstm_ctc_v2.0_train.tar)). It can be converted as follow:
``` ```
python3 tools/export_model.py -c configs/det/rec_r34_vd_none_bilstm_ctc.yml -o Global.pretrained_model=./rec_r34_vd_none_bilstm_ctc_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/rec_crnn python3 tools/export_model.py -c configs/det/rec_r34_vd_none_bilstm_ctc.yml -o Global.pretrained_model=./rec_r34_vd_none_bilstm_ctc_v2.0_train/best_accuracy Global.save_inference_dir=./inference/rec_crnn
``` ```
For CRNN text recognition model inference, execute the following commands: For CRNN text recognition model inference, execute the following commands:
......
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