营业执照识别

最新推荐文章于 2024-05-16 09:31:32 发布

ronshi

最新推荐文章于 2024-05-16 09:31:32 发布

阅读量462

点赞数

分类专栏： AI 文章标签：深度学习机器学习人工智能

本文链接：https://blog.csdn.net/ronshi/article/details/134446737

版权

AI 专栏收录该内容

6 篇文章 0 订阅

订阅专栏

1、制作文本检测数据集

文本检测数据集格式如下：

 
dataset_dir     # 数据集根目录，目录名称可以改变
├── images      # 存放图像的目录，目录名称可以改变，但要注意和train.txt val.txt的内容对应
├── train.txt   # 训练集标注文件，文件名称不可改变，内容举例：images/img_0.jpg 	 [{"transcription": "张三公司", "points": [[310, 104], [416, 141], [418, 216], [312, 179]]}, {...}]
└── val.txt     # 验证集标注文件，文件名称不可改变，内容举例：images/img_61.jpg 	 [{"transcription": "李四公司", "points": [[31, 10], [310, 140], [420, 220], [310, 170]]}, {...}]

2、导出文本检测推理模型

文本检测推理模型结构如下：

3、制作语义实体识别数据集

语义实体识别数据集格式如下：

train_data/XFUND    
├── zh_train    
|      ├── images
|      └── train.json 
├── zh_val
|      ├── images  
|      └── val.json
└── class_list_xfun.txt

4、训练模型

python tools/train.py -c configs/kie/vi_layoutxlm/ser_vi_layoutxlm_xfund_zh_udml.yml

配置文件 ser_vi_layoutxlm_xfund_zh_udml.yml

Global:
  use_gpu: True
  epoch_num: &epoch_num 200
  log_smooth_window: 10
  print_batch_step: 10
  save_model_dir: ./output/ser_vi_layoutxlm_xfund_zh_udml
  save_epoch_step: 2000
  # evaluation is run every 10 iterations after the 0th iteration
  eval_batch_step: [ 0, 19 ]
  cal_metric_during_train: False
  save_inference_dir:
  use_visualdl: False
  seed: 2022
  infer_img: ppstructure/docs/kie/input/zh_val_42.jpg
  save_res_path: ./output/ser_layoutxlm_xfund_zh/res


Architecture:
  model_type: &model_type "kie"
  name: DistillationModel
  algorithm: Distillation
  Models:
    Teacher:
      pretrained:
      freeze_params: false
      return_all_feats: true
      model_type: *model_type
      algorithm: &algorithm "LayoutXLM"
      Transform:
      Backbone:
        name: LayoutXLMForSer
        pretrained: True
        # one of base or vi
        mode: vi
        checkpoints:
        num_classes: &num_classes 5
    Student:
      pretrained:
      freeze_params: false
      return_all_feats: true
      model_type: *model_type
      algorithm: *algorithm
      Transform:
      Backbone:
        name: LayoutXLMForSer
        pretrained: True
        # one of base or vi
        mode: vi
        checkpoints:
        num_classes: *num_classes


Loss:
  name: CombinedLoss
  loss_config_list:
  - DistillationVQASerTokenLayoutLMLoss:
      weight: 1.0
      model_name_list: ["Student", "Teacher"]
      key: backbone_out
      num_classes: *num_classes
  - DistillationSERDMLLoss:
      weight: 1.0
      act: "softmax"
      use_log: true
      model_name_pairs:
      - ["Student", "Teacher"]
      key: backbone_out
  - DistillationVQADistanceLoss:
      weight: 0.5
      mode: "l2"
      model_name_pairs:
        - ["Student", "Teacher"]
      key: hidden_states_5
      name: "loss_5"
  - DistillationVQADistanceLoss:
      weight: 0.5
      mode: "l2"
      model_name_pairs:
        - ["Student", "Teacher"]
      key: hidden_states_8
      name: "loss_8"
  
  

Optimizer:
  name: AdamW
  beta1: 0.9
  beta2: 0.999
  lr:
    name: Linear
    learning_rate: 0.00005
    epochs: *epoch_num
    warmup_epoch: 10
  regularizer:
    name: L2
    factor: 0.00000
    
PostProcess:
  name: DistillationSerPostProcess
  model_name: ["Student", "Teacher"]
  key: backbone_out
  class_path: &class_path train_data/XFUND/class_list_xfun.txt

Metric:
  name: DistillationMetric
  base_metric_name: VQASerTokenMetric
  main_indicator: hmean
  key: "Student"

Train:
  dataset:
    name: SimpleDataSet
    data_dir: train_data/XFUND/zh_train
    label_file_list: 
      - train_data/XFUND/zh_train/train.json
    ratio_list: [ 1.0 ]
    transforms:
      - DecodeImage: # load image
          img_mode: RGB
          channel_first: False
      - VQATokenLabelEncode: # Class handling label
          contains_re: False
          algorithm: *algorithm
          class_path: *class_path
          # one of [None, "tb-yx"]
          order_method: &order_method "tb-yx"
      - VQATokenPad:
          max_seq_len: &max_seq_len 512
          return_attention_mask: True
      - VQASerTokenChunk:
          max_seq_len: *max_seq_len
      - Resize:
          size: [224,224]
      - NormalizeImage:
          scale: 1
          mean: [ 123.675, 116.28, 103.53 ]
          std: [ 58.395, 57.12, 57.375 ]
          order: 'hwc'
      - ToCHWImage:
      - KeepKeys:
          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels'] # dataloader will return list in this order
  loader:
    shuffle: True
    drop_last: False
    batch_size_per_card: 1
    num_workers: 1

Eval:
  dataset:
    name: SimpleDataSet
    data_dir: train_data/XFUND/zh_val
    label_file_list:
      - train_data/XFUND/zh_val/val.json
    transforms:
      - DecodeImage: # load image
          img_mode: RGB
          channel_first: False
      - VQATokenLabelEncode: # Class handling label
          contains_re: False
          algorithm: *algorithm
          class_path: *class_path
          order_method: *order_method
      - VQATokenPad:
          max_seq_len: *max_seq_len
          return_attention_mask: True
      - VQASerTokenChunk:
          max_seq_len: *max_seq_len
      - Resize:
          size: [224,224]
      - NormalizeImage:
          scale: 1
          mean: [ 123.675, 116.28, 103.53 ]
          std: [ 58.395, 57.12, 57.375 ]
          order: 'hwc'
      - ToCHWImage:
      - KeepKeys:
          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels'] # dataloader will return list in this order
  loader:
    shuffle: False
    drop_last: False
    batch_size_per_card: 8
    num_workers: 4

5、模型预测

python tools/infer_kie_token_ser.py \
  -c configs/kie/vi_layoutxlm/ser_vi_layoutxlm_xfund_zh.yml \
  -o Architecture.Backbone.checkpoints=./output/ser_vi_layoutxlm_xfund_zh_udml/best_accuracy \
  Global.infer_img=./train_data/XFUND/zh_train/images/51.jpg  Global.kie_det_model_dir="./infer_model/inference"

配置文件 ser_vi_layoutxlm_xfund_zh.yml

Global:
  use_gpu: True
  epoch_num: &epoch_num 200
  log_smooth_window: 10
  print_batch_step: 10
  save_model_dir: ./output/ser_vi_layoutxlm_xfund_zh
  save_epoch_step: 2000
  # evaluation is run every 10 iterations after the 0th iteration
  eval_batch_step: [ 0, 19 ]
  cal_metric_during_train: False
  save_inference_dir:
  use_visualdl: False
  seed: 2022
  infer_img: ppstructure/docs/kie/input/zh_val_42.jpg
  d2s_train_image_shape: [3, 224, 224]
  # if you want to predict using the groundtruth ocr info,
  # you can use the following config
  # infer_img: train_data/XFUND/zh_val/val.json
  # infer_mode: False

  save_res_path: ./output/ser/xfund_zh/res
  kie_rec_model_dir: 
  kie_det_model_dir:
  amp_custom_white_list: ['scale', 'concat', 'elementwise_add']

Architecture:
  model_type: kie
  algorithm: &algorithm "LayoutXLM"
  Transform:
  Backbone:
    name: LayoutXLMForSer
    pretrained: True
    checkpoints:
    # one of base or vi
    mode: vi
    num_classes: &num_classes 3

Loss:
  name: VQASerTokenLayoutLMLoss
  num_classes: *num_classes
  key: "backbone_out"

Optimizer:
  name: AdamW
  beta1: 0.9
  beta2: 0.999
  lr:
    name: Linear
    learning_rate: 0.00005
    epochs: *epoch_num
    warmup_epoch: 2
  regularizer:
    name: L2
    factor: 0.00000
    
PostProcess:
  name: VQASerTokenLayoutLMPostProcess
  class_path: &class_path train_data/XFUND/class_list_xfun.txt

Metric:
  name: VQASerTokenMetric
  main_indicator: hmean

Train:
  dataset:
    name: SimpleDataSet
    data_dir: train_data/XFUND/zh_train
    label_file_list: 
      - train_data/XFUND/zh_train/train.json
    ratio_list: [ 1.0 ]
    transforms:
      - DecodeImage: # load image
          img_mode: RGB
          channel_first: False
      - VQATokenLabelEncode: # Class handling label
          contains_re: False
          algorithm: *algorithm
          class_path: *class_path
          use_textline_bbox_info: &use_textline_bbox_info True
          # one of [None, "tb-yx"]
          order_method: &order_method "tb-yx"
      - VQATokenPad:
          max_seq_len: &max_seq_len 512
          return_attention_mask: True
      - VQASerTokenChunk:
          max_seq_len: *max_seq_len
      - Resize:
          size: [224,224]
      - NormalizeImage:
          scale: 1
          mean: [ 123.675, 116.28, 103.53 ]
          std: [ 58.395, 57.12, 57.375 ]
          order: 'hwc'
      - ToCHWImage:
      - KeepKeys:
          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels'] # dataloader will return list in this order
  loader:
    shuffle: True
    drop_last: False
    batch_size_per_card: 8
    num_workers: 4

Eval:
  dataset:
    name: SimpleDataSet
    data_dir: train_data/XFUND/zh_val
    label_file_list:
      - train_data/XFUND/zh_val/val.json
    transforms:
      - DecodeImage: # load image
          img_mode: RGB
          channel_first: False
      - VQATokenLabelEncode: # Class handling label
          contains_re: False
          algorithm: *algorithm
          class_path: *class_path
          use_textline_bbox_info: *use_textline_bbox_info
          order_method: *order_method
      - VQATokenPad:
          max_seq_len: *max_seq_len
          return_attention_mask: True
      - VQASerTokenChunk:
          max_seq_len: *max_seq_len
      - Resize:
          size: [224,224]
      - NormalizeImage:
          scale: 1
          mean: [ 123.675, 116.28, 103.53 ]
          std: [ 58.395, 57.12, 57.375 ]
          order: 'hwc'
      - ToCHWImage:
      - KeepKeys:
          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels'] # dataloader will return list in this order
  loader:
    shuffle: False
    drop_last: False
    batch_size_per_card: 8
    num_workers: 4

6、模型导出

python tools/export_model.py -c configs/kie/vi_layoutxlm/ser_vi_layoutxlm_xfund_zh.yml -o Architecture.Backbone.checkpoints=./output/ser_vi_layoutxlm_xfund_zh_udml/best_accuracy Global.kie_det_model_dir="./infer_model/inference" Global.save_inference_dir=./inference/ser_vi_layoutxlm

7、模型推理

python ppstructure/kie/predict_kie_token_ser.py \
  --kie_algorithm=LayoutXLM \
  --ser_model_dir=./inference/ser_vi_layoutxlm \
  --det_model_dir=./infer_model/inference \
  --image_dir=./train_data/XFUND/zh_train/images/18.jpg \
  --ser_dict_path=./train_data/XFUND/class_list_xfun.txt \
  --vis_font_path=./doc/fonts/simfang.ttf \
  --ocr_order_method="tb-yx"