infer.py 22.0 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14
# 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.

G
Guanghua Yu 已提交
15 16 17 18 19 20 21 22 23 24
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import os, sys
# add python path of PadleDetection to sys.path
parent_path = os.path.abspath(os.path.join(__file__, *(['..'] * 3)))
if parent_path not in sys.path:
    sys.path.append(parent_path)

25
import argparse
26
import time
27
import yaml
C
channings 已提交
28 29
import ast
from functools import reduce
30

31 32 33
from PIL import Image
import cv2
import numpy as np
34
import paddle
35
import paddle.fluid as fluid
G
Guanghua Yu 已提交
36
from preprocess import preprocess, Resize, Normalize, Permute, PadStride
37
from visualize import visualize_box_mask, lmk2out
38

39 40 41 42 43 44 45 46 47 48
# Global dictionary
SUPPORT_MODELS = {
    'YOLO',
    'SSD',
    'RetinaNet',
    'EfficientDet',
    'RCNN',
    'Face',
    'TTF',
    'FCOS',
G
Guanghua Yu 已提交
49
    'SOLOv2',
50 51
}

52

53
class Detector(object):
54 55
    """
    Args:
G
Guanghua Yu 已提交
56 57 58 59 60
        config (object): config of model, defined by `Config(model_dir)`
        model_dir (str): root path of __model__, __params__ and infer_cfg.yml
        use_gpu (bool): whether use gpu
        run_mode (str): mode of running(fluid/trt_fp32/trt_fp16)
        threshold (float): threshold to reserve the result for output.
61 62 63
    """

    def __init__(self,
G
Guanghua Yu 已提交
64 65 66 67 68 69 70 71 72
                 config,
                 model_dir,
                 use_gpu=False,
                 run_mode='fluid',
                 threshold=0.5):
        self.config = config
        if self.config.use_python_inference:
            self.executor, self.program, self.fecth_targets = load_executor(
                model_dir, use_gpu=use_gpu)
73
        else:
G
Guanghua Yu 已提交
74 75 76 77 78
            self.predictor = load_predictor(
                model_dir,
                run_mode=run_mode,
                min_subgraph_size=self.config.min_subgraph_size,
                use_gpu=use_gpu)
79

G
Guanghua Yu 已提交
80 81 82
    def preprocess(self, im):
        preprocess_ops = []
        for op_info in self.config.preprocess_infos:
83 84
            new_op_info = op_info.copy()
            op_type = new_op_info.pop('type')
G
Guanghua Yu 已提交
85
            if op_type == 'Resize':
86 87
                new_op_info['arch'] = self.config.arch
            preprocess_ops.append(eval(op_type)(**new_op_info))
G
Guanghua Yu 已提交
88 89 90
        im, im_info = preprocess(im, preprocess_ops)
        inputs = create_inputs(im, im_info, self.config.arch)
        return inputs, im_info
91

92
    def postprocess(self, np_boxes, np_masks, np_lmk, im_info, threshold=0.5):
G
Guanghua Yu 已提交
93 94
        # postprocess output of predictor
        results = {}
95 96 97
        if np_lmk is not None:
            results['landmark'] = lmk2out(np_boxes, np_lmk, im_info, threshold)

G
Guanghua Yu 已提交
98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115
        if self.config.arch in ['SSD', 'Face']:
            w, h = im_info['origin_shape']
            np_boxes[:, 2] *= h
            np_boxes[:, 3] *= w
            np_boxes[:, 4] *= h
            np_boxes[:, 5] *= w
        expect_boxes = (np_boxes[:, 1] > threshold) & (np_boxes[:, 0] > -1)
        np_boxes = np_boxes[expect_boxes, :]
        for box in np_boxes:
            print('class_id:{:d}, confidence:{:.4f},'
                  'left_top:[{:.2f},{:.2f}],'
                  ' right_bottom:[{:.2f},{:.2f}]'.format(
                      int(box[0]), box[1], box[2], box[3], box[4], box[5]))
        results['boxes'] = np_boxes
        if np_masks is not None:
            np_masks = np_masks[expect_boxes, :, :, :]
            results['masks'] = np_masks
        return results
116

G
Guanghua Yu 已提交
117 118 119 120 121 122 123
    def predict(self,
                image,
                threshold=0.5,
                warmup=0,
                repeats=1,
                run_benchmark=False):
        '''
124
        Args:
G
Guanghua Yu 已提交
125 126
            image (str/np.ndarray): path of image/ np.ndarray read by cv2
            threshold (float): threshold of predicted box' score
127
        Returns:
G
Guanghua Yu 已提交
128 129 130 131 132 133
            results (dict): include 'boxes': np.ndarray: shape:[N,6], N: number of box,
                            matix element:[class, score, x_min, y_min, x_max, y_max]
                            MaskRCNN's results include 'masks': np.ndarray:
                            shape:[N, class_num, mask_resolution, mask_resolution]
        '''
        inputs, im_info = self.preprocess(image)
134
        np_boxes, np_masks, np_lmk = None, None, None
G
Guanghua Yu 已提交
135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152
        if self.config.use_python_inference:
            for i in range(warmup):
                outs = self.executor.run(self.program,
                                         feed=inputs,
                                         fetch_list=self.fecth_targets,
                                         return_numpy=False)
            t1 = time.time()
            for i in range(repeats):
                outs = self.executor.run(self.program,
                                         feed=inputs,
                                         fetch_list=self.fecth_targets,
                                         return_numpy=False)
            t2 = time.time()
            ms = (t2 - t1) * 1000.0 / repeats
            print("Inference: {} ms per batch image".format(ms))
            np_boxes = np.array(outs[0])
            if self.config.mask_resolution is not None:
                np_masks = np.array(outs[1])
153
        else:
G
Guanghua Yu 已提交
154 155 156 157
            input_names = self.predictor.get_input_names()
            for i in range(len(input_names)):
                input_tensor = self.predictor.get_input_tensor(input_names[i])
                input_tensor.copy_from_cpu(inputs[input_names[i]])
158

G
Guanghua Yu 已提交
159 160 161 162 163 164 165 166 167
            for i in range(warmup):
                self.predictor.zero_copy_run()
                output_names = self.predictor.get_output_names()
                boxes_tensor = self.predictor.get_output_tensor(output_names[0])
                np_boxes = boxes_tensor.copy_to_cpu()
                if self.config.mask_resolution is not None:
                    masks_tensor = self.predictor.get_output_tensor(
                        output_names[1])
                    np_masks = masks_tensor.copy_to_cpu()
168

169 170 171 172 173 174 175 176 177 178 179
                if self.config.with_lmk is not None and self.config.with_lmk == True:
                    face_index = self.predictor.get_output_tensor(output_names[
                        1])
                    landmark = self.predictor.get_output_tensor(output_names[2])
                    prior_boxes = self.predictor.get_output_tensor(output_names[
                        3])
                    np_face_index = face_index.copy_to_cpu()
                    np_prior_boxes = prior_boxes.copy_to_cpu()
                    np_landmark = landmark.copy_to_cpu()
                    np_lmk = [np_face_index, np_landmark, np_prior_boxes]

G
Guanghua Yu 已提交
180 181 182 183 184 185 186 187 188 189
            t1 = time.time()
            for i in range(repeats):
                self.predictor.zero_copy_run()
                output_names = self.predictor.get_output_names()
                boxes_tensor = self.predictor.get_output_tensor(output_names[0])
                np_boxes = boxes_tensor.copy_to_cpu()
                if self.config.mask_resolution is not None:
                    masks_tensor = self.predictor.get_output_tensor(
                        output_names[1])
                    np_masks = masks_tensor.copy_to_cpu()
190 191 192 193 194 195 196 197 198 199 200

                if self.config.with_lmk is not None and self.config.with_lmk == True:
                    face_index = self.predictor.get_output_tensor(output_names[
                        1])
                    landmark = self.predictor.get_output_tensor(output_names[2])
                    prior_boxes = self.predictor.get_output_tensor(output_names[
                        3])
                    np_face_index = face_index.copy_to_cpu()
                    np_prior_boxes = prior_boxes.copy_to_cpu()
                    np_landmark = landmark.copy_to_cpu()
                    np_lmk = [np_face_index, np_landmark, np_prior_boxes]
G
Guanghua Yu 已提交
201 202 203
            t2 = time.time()
            ms = (t2 - t1) * 1000.0 / repeats
            print("Inference: {} ms per batch image".format(ms))
204

G
Guanghua Yu 已提交
205 206 207 208 209 210 211 212
        # do not perform postprocess in benchmark mode
        results = []
        if not run_benchmark:
            if reduce(lambda x, y: x * y, np_boxes.shape) < 6:
                print('[WARNNING] No object detected.')
                results = {'boxes': np.array([])}
            else:
                results = self.postprocess(
213
                    np_boxes, np_masks, np_lmk, im_info, threshold=threshold)
214

G
Guanghua Yu 已提交
215
        return results
216 217


G
Guanghua Yu 已提交
218 219 220 221 222 223 224 225 226 227 228 229 230
class DetectorSOLOv2(Detector):
    def __init__(self,
                 config,
                 model_dir,
                 use_gpu=False,
                 run_mode='fluid',
                 threshold=0.5):
        super(DetectorSOLOv2, self).__init__(
            config=config,
            model_dir=model_dir,
            use_gpu=use_gpu,
            run_mode=run_mode,
            threshold=threshold)
231

G
Guanghua Yu 已提交
232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270
    def predict(self,
                image,
                threshold=0.5,
                warmup=0,
                repeats=1,
                run_benchmark=False):
        inputs, im_info = self.preprocess(image)
        np_label, np_score, np_segms = None, None, None
        if self.config.use_python_inference:
            for i in range(warmup):
                outs = self.executor.run(self.program,
                                         feed=inputs,
                                         fetch_list=self.fecth_targets,
                                         return_numpy=False)
            t1 = time.time()
            for i in range(repeats):
                outs = self.executor.run(self.program,
                                         feed=inputs,
                                         fetch_list=self.fecth_targets,
                                         return_numpy=False)
            t2 = time.time()
            ms = (t2 - t1) * 1000.0 / repeats
            print("Inference: {} ms per batch image".format(ms))
            np_label, np_score, np_segms = np.array(outs[0]), np.array(outs[
                1]), np.array(outs[2])
        else:
            input_names = self.predictor.get_input_names()
            for i in range(len(input_names)):
                input_tensor = self.predictor.get_input_tensor(input_names[i])
                input_tensor.copy_from_cpu(inputs[input_names[i]])
            for i in range(warmup):
                self.predictor.zero_copy_run()
                output_names = self.predictor.get_output_names()
                np_label = self.predictor.get_output_tensor(output_names[
                    0]).copy_to_cpu()
                np_score = self.predictor.get_output_tensor(output_names[
                    1]).copy_to_cpu()
                np_segms = self.predictor.get_output_tensor(output_names[
                    2]).copy_to_cpu()
271

G
Guanghua Yu 已提交
272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290
            t1 = time.time()
            for i in range(repeats):
                self.predictor.zero_copy_run()
                output_names = self.predictor.get_output_names()
                np_label = self.predictor.get_output_tensor(output_names[
                    0]).copy_to_cpu()
                np_score = self.predictor.get_output_tensor(output_names[
                    1]).copy_to_cpu()
                np_segms = self.predictor.get_output_tensor(output_names[
                    2]).copy_to_cpu()
            t2 = time.time()
            ms = (t2 - t1) * 1000.0 / repeats
            print("Inference: {} ms per batch image".format(ms))

        # do not perform postprocess in benchmark mode
        results = []
        if not run_benchmark:
            return dict(segm=np_segms, label=np_label, score=np_score)
        return results
291 292 293 294 295 296 297 298 299 300 301 302 303 304 305


def create_inputs(im, im_info, model_arch='YOLO'):
    """generate input for different model type
    Args:
        im (np.ndarray): image (np.ndarray)
        im_info (dict): info of image
        model_arch (str): model type
    Returns:
        inputs (dict): input of model
    """
    inputs = {}
    inputs['image'] = im
    origin_shape = list(im_info['origin_shape'])
    resize_shape = list(im_info['resize_shape'])
G
Guanghua Yu 已提交
306 307
    pad_shape = list(im_info['pad_shape']) if im_info[
        'pad_shape'] is not None else list(im_info['resize_shape'])
W
wangguanzhong 已提交
308
    scale_x, scale_y = im_info['scale']
309 310 311
    if 'YOLO' in model_arch:
        im_size = np.array([origin_shape]).astype('int32')
        inputs['im_size'] = im_size
312
    elif 'RetinaNet' in model_arch or 'EfficientDet' in model_arch:
W
wangguanzhong 已提交
313
        scale = scale_x
G
Guanghua Yu 已提交
314
        im_info = np.array([pad_shape + [scale]]).astype('float32')
315
        inputs['im_info'] = im_info
316
    elif ('RCNN' in model_arch) or ('FCOS' in model_arch):
W
wangguanzhong 已提交
317
        scale = scale_x
G
Guanghua Yu 已提交
318
        im_info = np.array([pad_shape + [scale]]).astype('float32')
319 320 321
        im_shape = np.array([origin_shape + [1.]]).astype('float32')
        inputs['im_info'] = im_info
        inputs['im_shape'] = im_shape
W
wangguanzhong 已提交
322 323 324
    elif 'TTF' in model_arch:
        scale_factor = np.array([scale_x, scale_y] * 2).astype('float32')
        inputs['scale_factor'] = scale_factor
G
Guanghua Yu 已提交
325 326 327 328
    elif 'SOLOv2' in model_arch:
        scale = scale_x
        im_info = np.array([resize_shape + [scale]]).astype('float32')
        inputs['im_info'] = im_info
329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351
    return inputs


class Config():
    """set config of preprocess, postprocess and visualize
    Args:
        model_dir (str): root path of model.yml
    """

    def __init__(self, model_dir):
        # parsing Yaml config for Preprocess
        deploy_file = os.path.join(model_dir, 'infer_cfg.yml')
        with open(deploy_file) as f:
            yml_conf = yaml.safe_load(f)
        self.check_model(yml_conf)
        self.arch = yml_conf['arch']
        self.preprocess_infos = yml_conf['Preprocess']
        self.use_python_inference = yml_conf['use_python_inference']
        self.min_subgraph_size = yml_conf['min_subgraph_size']
        self.labels = yml_conf['label_list']
        self.mask_resolution = None
        if 'mask_resolution' in yml_conf:
            self.mask_resolution = yml_conf['mask_resolution']
352 353 354
        self.with_lmk = None
        if 'with_lmk' in yml_conf:
            self.with_lmk = yml_conf['with_lmk']
C
channings 已提交
355
        self.print_config()
356 357 358 359 360 361

    def check_model(self, yml_conf):
        """
        Raises:
            ValueError: loaded model not in supported model type 
        """
362
        for support_model in SUPPORT_MODELS:
363 364
            if support_model in yml_conf['arch']:
                return True
W
wangguanzhong 已提交
365
        raise ValueError("Unsupported arch: {}, expect {}".format(yml_conf[
366
            'arch'], SUPPORT_MODELS))
367

C
channings 已提交
368 369 370
    def print_config(self):
        print('-----------  Model Configuration -----------')
        print('%s: %s' % ('Model Arch', self.arch))
371
        print('%s: %s' % ('Use Paddle Executor', self.use_python_inference))
C
channings 已提交
372 373 374 375 376
        print('%s: ' % ('Transform Order'))
        for op_info in self.preprocess_infos:
            print('--%s: %s' % ('transform op', op_info['type']))
        print('--------------------------------------------')

377 378 379 380 381 382

def load_predictor(model_dir,
                   run_mode='fluid',
                   batch_size=1,
                   use_gpu=False,
                   min_subgraph_size=3):
383
    """set AnalysisConfig, generate AnalysisPredictor
384 385 386 387 388 389
    Args:
        model_dir (str): root path of __model__ and __params__
        use_gpu (bool): whether use gpu
    Returns:
        predictor (PaddlePredictor): AnalysisPredictor
    Raises:
390
        ValueError: predict by TensorRT need use_gpu == True.
391
    """
392
    if not use_gpu and not run_mode == 'fluid':
393 394 395
        raise ValueError(
            "Predict by TensorRT mode: {}, expect use_gpu==True, but use_gpu == {}"
            .format(run_mode, use_gpu))
396
    use_calib_mode = True if run_mode == 'trt_int8' else False
397
    precision_map = {
C
channings 已提交
398
        'trt_int8': fluid.core.AnalysisConfig.Precision.Int8,
399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414
        'trt_fp32': fluid.core.AnalysisConfig.Precision.Float32,
        'trt_fp16': fluid.core.AnalysisConfig.Precision.Half
    }
    config = fluid.core.AnalysisConfig(
        os.path.join(model_dir, '__model__'),
        os.path.join(model_dir, '__params__'))
    if use_gpu:
        # initial GPU memory(M), device ID
        config.enable_use_gpu(100, 0)
        # optimize graph and fuse op
        config.switch_ir_optim(True)
    else:
        config.disable_gpu()

    if run_mode in precision_map.keys():
        config.enable_tensorrt_engine(
415
            workspace_size=1 << 10,
416 417 418 419
            max_batch_size=batch_size,
            min_subgraph_size=min_subgraph_size,
            precision_mode=precision_map[run_mode],
            use_static=False,
420
            use_calib_mode=use_calib_mode)
421 422 423 424 425

    # disable print log when predict
    config.disable_glog_info()
    # enable shared memory
    config.enable_memory_optim()
426
    # disable feed, fetch OP, needed by zero_copy_run
427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449
    config.switch_use_feed_fetch_ops(False)
    predictor = fluid.core.create_paddle_predictor(config)
    return predictor


def load_executor(model_dir, use_gpu=False):
    if use_gpu:
        place = fluid.CUDAPlace(0)
    else:
        place = fluid.CPUPlace()
    exe = fluid.Executor(place)
    program, feed_names, fetch_targets = fluid.io.load_inference_model(
        dirname=model_dir,
        executor=exe,
        model_filename='__model__',
        params_filename='__params__')
    return exe, program, fetch_targets


def visualize(image_file,
              results,
              labels,
              mask_resolution=14,
G
Guanghua Yu 已提交
450 451
              output_dir='output/',
              threshold=0.5):
452 453
    # visualize the predict result
    im = visualize_box_mask(
G
Guanghua Yu 已提交
454 455 456 457 458
        image_file,
        results,
        labels,
        mask_resolution=mask_resolution,
        threshold=threshold)
459 460 461 462 463 464 465 466
    img_name = os.path.split(image_file)[-1]
    if not os.path.exists(output_dir):
        os.makedirs(output_dir)
    out_path = os.path.join(output_dir, img_name)
    im.save(out_path, quality=95)
    print("save result to: " + out_path)


G
Guanghua Yu 已提交
467 468 469 470 471
def print_arguments(args):
    print('-----------  Running Arguments -----------')
    for arg, value in sorted(vars(args).items()):
        print('%s: %s' % (arg, value))
    print('------------------------------------------')
472 473


G
Guanghua Yu 已提交
474
def predict_image(detector):
C
channings 已提交
475 476
    if FLAGS.run_benchmark:
        detector.predict(
K
Kaipeng Deng 已提交
477 478 479 480 481
            FLAGS.image_file,
            FLAGS.threshold,
            warmup=100,
            repeats=100,
            run_benchmark=True)
C
channings 已提交
482 483 484 485 486 487 488
    else:
        results = detector.predict(FLAGS.image_file, FLAGS.threshold)
        visualize(
            FLAGS.image_file,
            results,
            detector.config.labels,
            mask_resolution=detector.config.mask_resolution,
G
Guanghua Yu 已提交
489 490
            output_dir=FLAGS.output_dir,
            threshold=FLAGS.threshold)
491 492


G
Guanghua Yu 已提交
493
def predict_video(detector, camera_id):
C
channings 已提交
494 495 496 497 498 499
    if camera_id != -1:
        capture = cv2.VideoCapture(camera_id)
        video_name = 'output.mp4'
    else:
        capture = cv2.VideoCapture(FLAGS.video_file)
        video_name = os.path.split(FLAGS.video_file)[-1]
500 501 502
    fps = 30
    width = int(capture.get(cv2.CAP_PROP_FRAME_WIDTH))
    height = int(capture.get(cv2.CAP_PROP_FRAME_HEIGHT))
K
Kaipeng Deng 已提交
503
    fourcc = cv2.VideoWriter_fourcc(*'mp4v')
504
    if not os.path.exists(FLAGS.output_dir):
505
        os.makedirs(FLAGS.output_dir)
506 507 508 509 510 511 512 513 514 515 516 517 518 519
    out_path = os.path.join(FLAGS.output_dir, video_name)
    writer = cv2.VideoWriter(out_path, fourcc, fps, (width, height))
    index = 1
    while (1):
        ret, frame = capture.read()
        if not ret:
            break
        print('detect frame:%d' % (index))
        index += 1
        results = detector.predict(frame, FLAGS.threshold)
        im = visualize_box_mask(
            frame,
            results,
            detector.config.labels,
520 521
            mask_resolution=detector.config.mask_resolution,
            threshold=FLAGS.threshold)
522 523
        im = np.array(im)
        writer.write(im)
C
channings 已提交
524 525 526 527
        if camera_id != -1:
            cv2.imshow('Mask Detection', im)
            if cv2.waitKey(1) & 0xFF == ord('q'):
                break
528 529 530
    writer.release()


G
Guanghua Yu 已提交
531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546
def main():
    config = Config(FLAGS.model_dir)
    detector = Detector(
        config, FLAGS.model_dir, use_gpu=FLAGS.use_gpu, run_mode=FLAGS.run_mode)
    if config.arch == 'SOLOv2':
        detector = DetectorSOLOv2(
            config,
            FLAGS.model_dir,
            use_gpu=FLAGS.use_gpu,
            run_mode=FLAGS.run_mode)
    # predict from image
    if FLAGS.image_file != '':
        predict_image(detector)
    # predict from video file or camera video stream
    if FLAGS.video_file != '' or FLAGS.camera_id != -1:
        predict_video(detector, FLAGS.camera_id)
C
channings 已提交
547 548


549
if __name__ == '__main__':
550 551 552 553
    try:
        paddle.enable_static()
    except:
        pass
554 555 556 557 558 559
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument(
        "--model_dir",
        type=str,
        default=None,
        help=("Directory include:'__model__', '__params__', "
560
              "'infer_cfg.yml', created by tools/export_model.py."),
561 562 563 564 565
        required=True)
    parser.add_argument(
        "--image_file", type=str, default='', help="Path of image file.")
    parser.add_argument(
        "--video_file", type=str, default='', help="Path of video file.")
C
channings 已提交
566 567 568 569 570
    parser.add_argument(
        "--camera_id",
        type=int,
        default=-1,
        help="device id of camera to predict.")
571 572 573 574
    parser.add_argument(
        "--run_mode",
        type=str,
        default='fluid',
575
        help="mode of running(fluid/trt_fp32/trt_fp16/trt_int8)")
576
    parser.add_argument(
C
channings 已提交
577 578 579 580 581 582 583 584 585
        "--use_gpu",
        type=ast.literal_eval,
        default=False,
        help="Whether to predict with GPU.")
    parser.add_argument(
        "--run_benchmark",
        type=ast.literal_eval,
        default=False,
        help="Whether to predict a image_file repeatedly for benchmark")
586 587 588 589 590 591 592 593 594
    parser.add_argument(
        "--threshold", type=float, default=0.5, help="Threshold of score.")
    parser.add_argument(
        "--output_dir",
        type=str,
        default="output",
        help="Directory of output visualization files.")

    FLAGS = parser.parse_args()
C
channings 已提交
595
    print_arguments(FLAGS)
596 597
    if FLAGS.image_file != '' and FLAGS.video_file != '':
        assert "Cannot predict image and video at the same time"
G
Guanghua Yu 已提交
598 599

    main()