infer_rec.py 6.5 KB
Newer Older
L
LDOUBLEV 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
# 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 __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import numpy as np
L
LDOUBLEV 已提交
20 21 22 23 24
import os
import sys
__dir__ = os.path.dirname(__file__)
sys.path.append(__dir__)
sys.path.append(os.path.join(__dir__, '..'))
L
LDOUBLEV 已提交
25

T
tink2123 已提交
26

L
LDOUBLEV 已提交
27 28 29 30 31 32 33 34 35 36 37 38 39
def set_paddle_flags(**kwargs):
    for key, value in kwargs.items():
        if os.environ.get(key, None) is None:
            os.environ[key] = str(value)


# NOTE(paddle-dev): All of these flags should be
# set before `import paddle`. Otherwise, it would
# not take any effect.
set_paddle_flags(
    FLAGS_eager_delete_tensor_gb=0,  # enable GC to save memory
)

L
LDOUBLEV 已提交
40
import tools.program as program
L
LDOUBLEV 已提交
41 42 43 44 45 46 47
from paddle import fluid
from ppocr.utils.utility import initial_logger
logger = initial_logger()
from ppocr.data.reader_main import reader_main
from ppocr.utils.save_load import init_model
from ppocr.utils.character import CharacterOps
from ppocr.utils.utility import create_module
T
tink2123 已提交
48
from ppocr.utils.utility import get_image_file_list
L
LDOUBLEV 已提交
49 50 51 52 53 54 55


def main():
    config = program.load_config(FLAGS.config)
    program.merge_config(FLAGS.opt)
    logger.info(config)
    char_ops = CharacterOps(config['Global'])
T
tink2123 已提交
56
    loss_type = config['Global']['loss_type']
L
LDOUBLEV 已提交
57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78
    config['Global']['char_ops'] = char_ops

    # check if set use_gpu=True in paddlepaddle cpu version
    use_gpu = config['Global']['use_gpu']
    #     check_gpu(use_gpu)

    place = fluid.CUDAPlace(0) if use_gpu else fluid.CPUPlace()
    exe = fluid.Executor(place)

    rec_model = create_module(config['Architecture']['function'])(params=config)
    startup_prog = fluid.Program()
    eval_prog = fluid.Program()
    with fluid.program_guard(eval_prog, startup_prog):
        with fluid.unique_name.guard():
            _, outputs = rec_model(mode="test")
            fetch_name_list = list(outputs.keys())
            fetch_varname_list = [outputs[v].name for v in fetch_name_list]
    eval_prog = eval_prog.clone(for_test=True)
    exe.run(startup_prog)

    init_model(config, eval_prog, exe)

T
tink2123 已提交
79
    blobs = reader_main(config, 'test')()
T
tink2123 已提交
80
    infer_img = config['Global']['infer_img']
T
tink2123 已提交
81
    infer_list = get_image_file_list(infer_img)
T
tink2123 已提交
82 83 84 85
    max_img_num = len(infer_list)
    if len(infer_list) == 0:
        logger.info("Can not find img in infer_img dir.")
    for i in range(max_img_num):
littletomatodonkey's avatar
littletomatodonkey 已提交
86
        logger.info("infer_img:%s" % infer_list[i])
T
tink2123 已提交
87
        img = next(blobs)
T
tink2123 已提交
88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117
        if loss_type != "srn":
            predict = exe.run(program=eval_prog,
                              feed={"image": img},
                              fetch_list=fetch_varname_list,
                              return_numpy=False)
        else:
            encoder_word_pos_list = []
            gsrm_word_pos_list = []
            gsrm_slf_attn_bias1_list = []
            gsrm_slf_attn_bias2_list = []
            encoder_word_pos_list.append(img[1])
            gsrm_word_pos_list.append(img[2])
            gsrm_slf_attn_bias1_list.append(img[3])
            gsrm_slf_attn_bias2_list.append(img[4])

            encoder_word_pos_list = np.concatenate(
                encoder_word_pos_list, axis=0).astype(np.int64)
            gsrm_word_pos_list = np.concatenate(
                gsrm_word_pos_list, axis=0).astype(np.int64)
            gsrm_slf_attn_bias1_list = np.concatenate(
                gsrm_slf_attn_bias1_list, axis=0).astype(np.float32)
            gsrm_slf_attn_bias2_list = np.concatenate(
                gsrm_slf_attn_bias2_list, axis=0).astype(np.float32)

            predict = exe.run(program=eval_prog, \
                       feed={'image': img[0], 'encoder_word_pos': encoder_word_pos_list,
                             'gsrm_word_pos': gsrm_word_pos_list, 'gsrm_slf_attn_bias1': gsrm_slf_attn_bias1_list,
                             'gsrm_slf_attn_bias2': gsrm_slf_attn_bias2_list}, \
                       fetch_list=fetch_varname_list, \
                       return_numpy=False)
D
dyning 已提交
118 119
        if loss_type == "ctc":
            preds = np.array(predict[0])
L
LDOUBLEV 已提交
120 121 122
            preds = preds.reshape(-1)
            preds_lod = predict[0].lod()[0]
            preds_text = char_ops.decode(preds)
D
dyning 已提交
123 124 125 126
            probs = np.array(predict[1])
            ind = np.argmax(probs, axis=1)
            blank = probs.shape[1]
            valid_ind = np.where(ind != (blank - 1))[0]
L
fix bug  
LDOUBLEV 已提交
127
            if len(valid_ind) == 0:
128
                continue
D
dyning 已提交
129 130 131 132
            score = np.mean(probs[valid_ind, ind[valid_ind]])
        elif loss_type == "attention":
            preds = np.array(predict[0])
            probs = np.array(predict[1])
L
LDOUBLEV 已提交
133 134
            end_pos = np.where(preds[0, :] == 1)[0]
            if len(end_pos) <= 1:
D
dyning 已提交
135 136
                preds = preds[0, 1:]
                score = np.mean(probs[0, 1:])
L
LDOUBLEV 已提交
137
            else:
D
dyning 已提交
138 139 140 141
                preds = preds[0, 1:end_pos[1]]
                score = np.mean(probs[0, 1:end_pos[1]])
            preds = preds.reshape(-1)
            preds_text = char_ops.decode(preds)
T
tink2123 已提交
142 143 144 145 146 147 148 149 150 151 152 153
        elif loss_type == "srn":
            cur_pred = []
            preds = np.array(predict[0])
            preds = preds.reshape(-1)
            probs = np.array(predict[1])
            ind = np.argmax(probs, axis=1)
            valid_ind = np.where(preds != 37)[0]
            if len(valid_ind) == 0:
                continue
            score = np.mean(probs[valid_ind, ind[valid_ind]])
            preds = preds[:valid_ind[-1] + 1]
            preds_text = char_ops.decode(preds)
littletomatodonkey's avatar
littletomatodonkey 已提交
154 155 156
        logger.info("\t index: {}".format(preds))
        logger.info("\t word : {}".format(preds_text))
        logger.info("\t score: {}".format(score))
L
LDOUBLEV 已提交
157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176

    # save for inference model
    target_var = []
    for key, values in outputs.items():
        target_var.append(values)

    fluid.io.save_inference_model(
        "./output/",
        feeded_var_names=['image'],
        target_vars=target_var,
        executor=exe,
        main_program=eval_prog,
        model_filename="model",
        params_filename="params")


if __name__ == '__main__':
    parser = program.ArgsParser()
    FLAGS = parser.parse_args()
    main()