det_mv3_db.yml 3.3 KB
Newer Older
L
LDOUBLEV 已提交
1 2 3 4 5
Global:
  use_gpu: true
  epoch_num: 1200
  log_smooth_window: 20
  print_batch_step: 2
W
WenmuZhou 已提交
6
  save_model_dir: ./output/db_mv3/
W
WenmuZhou 已提交
7
  save_epoch_step: 1200
L
LDOUBLEV 已提交
8
  # evaluation is run every 5000 iterations after the 4000th iteration
D
dyning 已提交
9
  eval_batch_step: [4000, 5000]
W
WenmuZhou 已提交
10 11 12
  # if pretrained_model is saved in static mode, load_static_weights must set to True
  load_static_weights: True
  cal_metric_during_train: False
W
WenmuZhou 已提交
13
  pretrained_model: ./pretrain_models/MobileNetV3_large_x0_5_pretrained
D
dyning 已提交
14
  checkpoints: #./output/det_db_0.001_DiceLoss_256_pp_config_2.0b_4gpu/best_accuracy
L
LDOUBLEV 已提交
15
  save_inference_dir:
D
dyning 已提交
16
  use_visualdl: False
W
WenmuZhou 已提交
17 18
  infer_img: doc/imgs_en/img_10.jpg
  save_res_path: ./output/det_db/predicts_db.txt
L
LDOUBLEV 已提交
19

W
WenmuZhou 已提交
20
Architecture:
D
dyning 已提交
21
  model_type: det
W
WenmuZhou 已提交
22 23 24 25 26 27 28
  algorithm: DB
  Transform:
  Backbone:
    name: MobileNetV3
    scale: 0.5
    model_name: large
  Neck:
D
dyning 已提交
29
    name: DBFPN
W
WenmuZhou 已提交
30 31 32 33
    out_channels: 256
  Head:
    name: DBHead
    k: 50
L
LDOUBLEV 已提交
34 35

Loss:
W
WenmuZhou 已提交
36
  name: DBLoss
L
LDOUBLEV 已提交
37 38 39 40 41
  balance_loss: true
  main_loss_type: DiceLoss
  alpha: 5
  beta: 10
  ohem_ratio: 3
D
dyning 已提交
42 43 44 45 46 47 48 49 50 51 52 53
  
Optimizer:
  name: Adam
  beta1: 0.9
  beta2: 0.999
  learning_rate:
#    name: Cosine
    lr: 0.001
#    warmup_epoch: 0
  regularizer:
    name: 'L2'
    factor: 0
L
LDOUBLEV 已提交
54 55

PostProcess:
W
WenmuZhou 已提交
56
  name: DBPostProcess
L
LDOUBLEV 已提交
57
  thresh: 0.3
W
WenmuZhou 已提交
58
  box_thresh: 0.6
L
LDOUBLEV 已提交
59
  max_candidates: 1000
W
WenmuZhou 已提交
60 61 62 63 64 65
  unclip_ratio: 1.5

Metric:
  name: DetMetric
  main_indicator: hmean

D
dyning 已提交
66
Train:
W
WenmuZhou 已提交
67 68
  dataset:
    name: SimpleDataSet
D
dyning 已提交
69 70 71 72
    data_dir: ./train_data/icdar2015/text_localization/
    label_file_list:
      - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt
    ratio_list: [0.5]
W
WenmuZhou 已提交
73 74 75 76 77 78 79 80
    transforms:
      - DecodeImage: # load image
          img_mode: BGR
          channel_first: False
      - DetLabelEncode: # Class handling label
      - IaaAugment:
          augmenter_args:
            - { 'type': Fliplr, 'args': { 'p': 0.5 } }
D
dyning 已提交
81 82
            - { 'type': Affine, 'args': { 'rotate': [-10, 10] } }
            - { 'type': Resize, 'args': { 'size': [0.5, 3] } }
W
WenmuZhou 已提交
83
      - EastRandomCropData:
D
dyning 已提交
84
          size: [640, 640]
W
WenmuZhou 已提交
85 86 87 88 89 90 91 92 93 94 95
          max_tries: 50
          keep_ratio: true
      - MakeBorderMap:
          shrink_ratio: 0.4
          thresh_min: 0.3
          thresh_max: 0.7
      - MakeShrinkMap:
          shrink_ratio: 0.4
          min_text_size: 8
      - NormalizeImage:
          scale: 1./255.
D
dyning 已提交
96 97
          mean: [0.485, 0.456, 0.406]
          std: [0.229, 0.224, 0.225]
W
WenmuZhou 已提交
98 99
          order: 'hwc'
      - ToCHWImage:
D
dyning 已提交
100 101
      - KeepKeys:
          keep_keys: ['image', 'threshold_map', 'threshold_mask', 'shrink_map', 'shrink_mask'] # the order of the dataloader list
W
WenmuZhou 已提交
102 103 104
  loader:
    shuffle: True
    drop_last: False
D
dyning 已提交
105
    batch_size_per_card: 4
W
WenmuZhou 已提交
106
    num_workers: 8
W
WenmuZhou 已提交
107

D
dyning 已提交
108
Eval:
W
WenmuZhou 已提交
109 110
  dataset:
    name: SimpleDataSet
D
dyning 已提交
111 112 113
    data_dir: ./train_data/icdar2015/text_localization/
    label_file_list:
      - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt
W
WenmuZhou 已提交
114 115 116 117 118 119
    transforms:
      - DecodeImage: # load image
          img_mode: BGR
          channel_first: False
      - DetLabelEncode: # Class handling label
      - DetResizeForTest:
D
dyning 已提交
120
          image_shape: [736, 1280]
W
WenmuZhou 已提交
121 122
      - NormalizeImage:
          scale: 1./255.
D
dyning 已提交
123 124
          mean: [0.485, 0.456, 0.406]
          std: [0.229, 0.224, 0.225]
W
WenmuZhou 已提交
125 126
          order: 'hwc'
      - ToCHWImage:
D
dyning 已提交
127 128
      - KeepKeys:
          keep_keys: ['image', 'shape', 'polys', 'ignore_tags']
W
WenmuZhou 已提交
129 130 131
  loader:
    shuffle: False
    drop_last: False
D
dyning 已提交
132 133
    batch_size_per_card: 1 # must be 1
    num_workers: 2