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

README.md

@@ -72,6 +72,7 @@ PaddleOCR support a variety of cutting-edge algorithms related to OCR, and devel
 <img src="https://raw.githubusercontent.com/PaddlePaddle/PaddleOCR/dygraph/doc/joinus.PNG"  width = "200" height = "200" />
 </div>
 <a name="Supported-Chinese-model-list"></a>
+
 ## PP-OCR Series Model List（Update on September 8th）
 
 | Model introduction                                           | Model name                   | Recommended scene | Detection model                                              | Direction classifier                                         | Recognition model                                            |

README_ch.md

@@ -71,6 +71,8 @@ PaddleOCR旨在打造一套丰富、领先、且实用的OCR工具库，助力
 ## 《动手学OCR》电子书
 - [《动手学OCR》电子书📚](./doc/doc_ch/ocr_book.md)
 
+## 场景应用
+- PaddleOCR场景应用覆盖通用，制造、金融、交通行业的主要OCR垂类应用，在PP-OCR、PP-Structure的通用能力基础之上，以notebook的形式展示利用场景数据微调、模型优化方法、数据增广等内容，为开发者快速落地OCR应用提供示范与启发。详情可查看[README](./applications)。
 
 <a name="开源社区"></a>
 ## 开源社区

configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_cml.yml

@@ -28,7 +28,7 @@ Architecture:
       algorithm: DB
       Transform:
       Backbone:
-        name: ResNet
+        name: ResNet_vd
         layers: 18
       Neck:
         name: DBFPN

configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_distill.yml

@@ -45,7 +45,7 @@ Architecture:
       algorithm: DB
       Transform:
       Backbone:
-        name: ResNet
+        name: ResNet_vd
         layers: 18
       Neck:
         name: DBFPN

configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_dml.yml

@@ -65,7 +65,7 @@ Loss:
       - ["Student", "Teacher"]
       maps_name: "thrink_maps"
       weight: 1.0
-      act: "softmax"
+      # act: None
       model_name_pairs: ["Student", "Teacher"]
       key: maps
   - DistillationDBLoss:

configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml

@@ -61,7 +61,7 @@ Architecture:
       model_type: det
       algorithm: DB
       Backbone:
-        name: ResNet
+        name: ResNet_vd
         in_channels: 3
         layers: 50
       Neck:

configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_dml.yml

@@ -25,7 +25,7 @@ Architecture:
       model_type: det
       algorithm: DB
       Backbone:
-        name: ResNet
+        name: ResNet_vd
         in_channels: 3
         layers: 50
       Neck:
@@ -40,7 +40,7 @@ Architecture:
       model_type: det
       algorithm: DB
       Backbone:
-        name: ResNet
+        name: ResNet_vd
         in_channels: 3
         layers: 50
       Neck:
@@ -60,7 +60,7 @@ Loss:
       - ["Student", "Student2"]
       maps_name: "thrink_maps"
       weight: 1.0
-      act: "softmax"
+      # act: None
       model_name_pairs: ["Student", "Student2"]
       key: maps
   - DistillationDBLoss:

configs/det/ch_ppocr_v2.0/ch_det_res18_db_v2.0.yml

@@ -20,7 +20,7 @@ Architecture:
   algorithm: DB
   Transform:
   Backbone:
-    name: ResNet
+    name: ResNet_vd
     layers: 18
     disable_se: True
   Neck:

configs/det/det_r50_db++_icdar15.yml

+Global:
+  debug: false
+  use_gpu: true
+  epoch_num: 1000
+  log_smooth_window: 20
+  print_batch_step: 10
+  save_model_dir: ./output/det_r50_icdar15/
+  save_epoch_step: 200
+  eval_batch_step:
+  - 0
+  - 2000
+  cal_metric_during_train: false
+  pretrained_model: ./pretrain_models/ResNet50_dcn_asf_synthtext_pretrained
+  checkpoints: null
+  save_inference_dir: null
+  use_visualdl: false
+  infer_img: doc/imgs_en/img_10.jpg
+  save_res_path: ./checkpoints/det_db/predicts_db.txt
+Architecture:
+  model_type: det
+  algorithm: DB++
+  Transform: null
+  Backbone:
+    name: ResNet
+    layers: 50
+    dcn_stage: [False, True, True, True]
+  Neck:
+    name: DBFPN
+    out_channels: 256
+    use_asf: True
+  Head:
+    name: DBHead
+    k: 50
+Loss:
+  name: DBLoss
+  balance_loss: true
+  main_loss_type: BCELoss
+  alpha: 5
+  beta: 10
+  ohem_ratio: 3
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: DecayLearningRate
+    learning_rate: 0.007
+    epochs: 1000
+    factor: 0.9
+    end_lr: 0
+  weight_decay: 0.0001
+PostProcess:
+  name: DBPostProcess
+  thresh: 0.3
+  box_thresh: 0.6
+  max_candidates: 1000
+  unclip_ratio: 1.5
+Metric:
+  name: DetMetric
+  main_indicator: hmean
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/icdar2015/text_localization/                                                                        
+    label_file_list:                                                                                            
+    - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt                                                     
+    ratio_list:                                                                                                 
+    - 1.0 
+    transforms:
+    - DecodeImage:
+        img_mode: BGR
+        channel_first: false
+    - DetLabelEncode: null
+    - IaaAugment:
+        augmenter_args:
+        - type: Fliplr
+          args:
+            p: 0.5
+        - type: Affine
+          args:
+            rotate:
+            - -10
+            - 10
+        - type: Resize
+          args:
+            size:
+            - 0.5
+            - 3
+    - EastRandomCropData:
+        size:
+        - 640
+        - 640
+        max_tries: 10
+        keep_ratio: true
+    - MakeShrinkMap:
+        shrink_ratio: 0.4
+        min_text_size: 8
+    - MakeBorderMap:
+        shrink_ratio: 0.4
+        thresh_min: 0.3
+        thresh_max: 0.7
+    - NormalizeImage:
+        scale: 1./255.
+        mean:
+        - 0.48109378172549
+        - 0.45752457890196
+        - 0.40787054090196
+        std:
+        - 1.0
+        - 1.0
+        - 1.0
+        order: hwc
+    - ToCHWImage: null
+    - KeepKeys:
+        keep_keys:
+        - image
+        - threshold_map
+        - threshold_mask
+        - shrink_map
+        - shrink_mask
+  loader:
+    shuffle: true
+    drop_last: false
+    batch_size_per_card: 4
+    num_workers: 8
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/icdar2015/text_localization                                                                          
+    label_file_list:                                                                                            
+    - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt
+    transforms:
+    - DecodeImage:
+        img_mode: BGR
+        channel_first: false
+    - DetLabelEncode: null
+    - DetResizeForTest:
+        image_shape:
+        - 1152
+        - 2048
+    - NormalizeImage:
+        scale: 1./255.
+        mean:
+        - 0.48109378172549
+        - 0.45752457890196
+        - 0.40787054090196
+        std:
+        - 1.0
+        - 1.0
+        - 1.0
+        order: hwc
+    - ToCHWImage: null
+    - KeepKeys:
+        keep_keys:
+        - image
+        - shape
+        - polys
+        - ignore_tags
+  loader:
+    shuffle: false
+    drop_last: false
+    batch_size_per_card: 1
+    num_workers: 2
+profiler_options: null

configs/det/det_r50_db++_td_tr.yml

+Global:
+  debug: false
+  use_gpu: true
+  epoch_num: 1000
+  log_smooth_window: 20
+  print_batch_step: 10
+  save_model_dir: ./output/det_r50_td_tr/
+  save_epoch_step: 200
+  eval_batch_step:
+  - 0
+  - 2000
+  cal_metric_during_train: false
+  pretrained_model: ./pretrain_models/ResNet50_dcn_asf_synthtext_pretrained
+  checkpoints: null
+  save_inference_dir: null
+  use_visualdl: false
+  infer_img: doc/imgs_en/img_10.jpg
+  save_res_path: ./checkpoints/det_db/predicts_db.txt
+Architecture:
+  model_type: det
+  algorithm: DB++
+  Transform: null
+  Backbone:
+    name: ResNet
+    layers: 50
+    dcn_stage: [False, True, True, True]
+  Neck:
+    name: DBFPN
+    out_channels: 256
+    use_asf: True
+  Head:
+    name: DBHead
+    k: 50
+Loss:
+  name: DBLoss
+  balance_loss: true
+  main_loss_type: BCELoss
+  alpha: 5
+  beta: 10
+  ohem_ratio: 3
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: DecayLearningRate
+    learning_rate: 0.007
+    epochs: 1000
+    factor: 0.9
+    end_lr: 0
+  weight_decay: 0.0001
+PostProcess:
+  name: DBPostProcess
+  thresh: 0.3
+  box_thresh: 0.5
+  max_candidates: 1000
+  unclip_ratio: 1.5
+Metric:
+  name: DetMetric
+  main_indicator: hmean
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/                                                                                 
+    label_file_list:                                                                                            
+    - ./train_data/TD_TR/TD500/train_gt_labels.txt                                                           
+    - ./train_data/TD_TR/TR400/gt_labels.txt                                                                 
+    ratio_list:                                                                                                 
+    - 1.0                                                                                                       
+    - 1.0 
+    transforms:
+    - DecodeImage:
+        img_mode: BGR
+        channel_first: false
+    - DetLabelEncode: null
+    - IaaAugment:
+        augmenter_args:
+        - type: Fliplr
+          args:
+            p: 0.5
+        - type: Affine
+          args:
+            rotate:
+            - -10
+            - 10
+        - type: Resize
+          args:
+            size:
+            - 0.5
+            - 3
+    - EastRandomCropData:
+        size:
+        - 640
+        - 640
+        max_tries: 10
+        keep_ratio: true
+    - MakeShrinkMap:
+        shrink_ratio: 0.4
+        min_text_size: 8
+    - MakeBorderMap:
+        shrink_ratio: 0.4
+        thresh_min: 0.3
+        thresh_max: 0.7
+    - NormalizeImage:
+        scale: 1./255.
+        mean:
+        - 0.48109378172549
+        - 0.45752457890196
+        - 0.40787054090196
+        std:
+        - 1.0
+        - 1.0
+        - 1.0
+        order: hwc
+    - ToCHWImage: null
+    - KeepKeys:
+        keep_keys:
+        - image
+        - threshold_map
+        - threshold_mask
+        - shrink_map
+        - shrink_mask
+  loader:
+    shuffle: true
+    drop_last: false
+    batch_size_per_card: 4
+    num_workers: 8
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/                                                                          
+    label_file_list:                                                                                            
+    - ./train_data/TD_TR/TD500/test_gt_labels.txt
+    transforms:
+    - DecodeImage:
+        img_mode: BGR
+        channel_first: false
+    - DetLabelEncode: null
+    - DetResizeForTest:
+        image_shape:
+        - 736
+        - 736
+        keep_ratio: True
+    - NormalizeImage:
+        scale: 1./255.
+        mean:
+        - 0.48109378172549
+        - 0.45752457890196
+        - 0.40787054090196
+        std:
+        - 1.0
+        - 1.0
+        - 1.0
+        order: hwc
+    - ToCHWImage: null
+    - KeepKeys:
+        keep_keys:
+        - image
+        - shape
+        - polys
+        - ignore_tags
+  loader:
+    shuffle: false
+    drop_last: false
+    batch_size_per_card: 1
+    num_workers: 2
+profiler_options: null

configs/det/det_r50_vd_db.yml

@@ -20,7 +20,7 @@ Architecture:
   algorithm: DB
   Transform:
   Backbone:
-    name: ResNet
+    name: ResNet_vd
     layers: 50
   Neck:
     name: DBFPN

configs/det/det_r50_vd_dcn_fce_ctw.yml

@@ -21,7 +21,7 @@ Architecture:
   algorithm: FCE
   Transform:
   Backbone:
-    name: ResNet
+    name: ResNet_vd
     layers: 50
     dcn_stage: [False, True, True, True]
     out_indices: [1,2,3]

configs/det/det_r50_vd_east.yml

@@ -20,7 +20,7 @@ Architecture:
   algorithm: EAST
   Transform:
   Backbone:
-    name: ResNet
+    name: ResNet_vd
     layers: 50
   Neck:
     name: EASTFPN

configs/det/det_r50_vd_pse.yml

@@ -20,7 +20,7 @@ Architecture:
   algorithm: PSE
   Transform:
   Backbone:
-    name: ResNet
+    name: ResNet_vd
     layers: 50
   Neck:
     name: FPN

configs/det/det_res18_db_v2.0.yml

@@ -20,7 +20,7 @@ Architecture:
   algorithm: DB
   Transform:
   Backbone:
-    name: ResNet
+    name: ResNet_vd
     layers: 18
     disable_se: True
   Neck:

configs/kie/kie_unet_sdmgr.yml

@@ -17,7 +17,7 @@ Global:
   checkpoints:
   save_inference_dir:
   use_visualdl: False
-  class_path: ./train_data/wildreceipt/class_list.txt
+  class_path: &class_path ./train_data/wildreceipt/class_list.txt
   infer_img: ./train_data/wildreceipt/1.txt
   save_res_path: ./output/sdmgr_kie/predicts_kie.txt
   img_scale: [ 1024, 512 ]
@@ -72,6 +72,7 @@ Train:
           order: 'hwc'
       - KieLabelEncode: # Class handling label
           character_dict_path: ./train_data/wildreceipt/dict.txt
+          class_path: *class_path
       - KieResize:
       - ToCHWImage:
       - KeepKeys:
@@ -88,7 +89,6 @@ Eval:
     data_dir: ./train_data/wildreceipt
     label_file_list:
       - ./train_data/wildreceipt/wildreceipt_test.txt
-      # - /paddle/data/PaddleOCR/train_data/wildreceipt/1.txt
     transforms:
       - DecodeImage: # load image
           img_mode: RGB

configs/table/table_master.yml

+Global:
+  use_gpu: true
+  epoch_num: 17
+  log_smooth_window: 20
+  print_batch_step: 100
+  save_model_dir: ./output/table_master/
+  save_epoch_step: 17
+  eval_batch_step: [0,  6259]
+  cal_metric_during_train: true
+  pretrained_model: null
+  checkpoints: 
+  save_inference_dir: output/table_master/infer
+  use_visualdl: false
+  infer_img: ppstructure/docs/table/table.jpg
+  save_res_path: ./output/table_master
+  character_dict_path: ppocr/utils/dict/table_master_structure_dict.txt
+  infer_mode: false
+  max_text_length: 500
+  process_total_num: 0
+  process_cut_num: 0
+
+
+Optimizer:
+  name: Adam
+  beta1: 0.9
+  beta2: 0.999
+  lr:
+    name: MultiStepDecay
+    learning_rate: 0.001
+    milestones: [12, 15]
+    gamma: 0.1
+    warmup_epoch: 0.02
+  regularizer:
+    name: L2
+    factor: 0.0
+
+Architecture:
+  model_type: table
+  algorithm: TableMaster
+  Backbone:
+    name: TableResNetExtra
+    gcb_config:
+      ratio: 0.0625
+      headers: 1
+      att_scale: False
+      fusion_type: channel_add
+      layers: [False, True, True, True]
+    layers: [1,2,5,3]
+  Head:
+    name: TableMasterHead
+    hidden_size: 512
+    headers: 8
+    dropout: 0
+    d_ff: 2024
+    max_text_length: 500
+
+Loss:
+  name: TableMasterLoss
+  ignore_index: 42 # set to len of dict + 3
+
+PostProcess:
+  name: TableMasterLabelDecode
+  box_shape: pad
+
+Metric:
+  name: TableMetric
+  main_indicator: acc
+  compute_bbox_metric: False
+
+Train:
+  dataset:
+    name: PubTabDataSet
+    data_dir: train_data/table/pubtabnet/train/
+    label_file_list: [train_data/table/pubtabnet/PubTabNet_2.0.0_train.jsonl]
+    transforms:
+      - DecodeImage:
+          img_mode: BGR
+          channel_first: False
+      - TableMasterLabelEncode:
+          learn_empty_box: False
+          merge_no_span_structure: True
+          replace_empty_cell_token: True
+      - ResizeTableImage:
+          max_len: 480
+          resize_bboxes: True
+      - PaddingTableImage:
+          size: [480, 480]
+      - TableBoxEncode:
+          use_xywh: True
+      - NormalizeImage:
+          scale: 1./255.
+          mean: [0.5, 0.5, 0.5]
+          std: [0.5, 0.5, 0.5]
+          order: hwc
+      - ToCHWImage: null
+      - KeepKeys:
+          keep_keys: [image, structure, bboxes, bbox_masks, shape]
+  loader:
+    shuffle: True
+    batch_size_per_card: 10
+    drop_last: True
+    num_workers: 8
+
+Eval:
+  dataset:
+    name: PubTabDataSet
+    data_dir: train_data/table/pubtabnet/train/
+    label_file_list: [train_data/table/pubtabnet/PubTabNet_2.0.0_val.jsonl]
+    transforms:
+      - DecodeImage:
+          img_mode: BGR
+          channel_first: False
+      - TableMasterLabelEncode:
+          learn_empty_box: False
+          merge_no_span_structure: True
+          replace_empty_cell_token: True
+      - ResizeTableImage:
+          max_len: 480
+          resize_bboxes: True
+      - PaddingTableImage:
+          size: [480, 480]
+      - TableBoxEncode:
+          use_xywh: True
+      - NormalizeImage:
+          scale: 1./255.
+          mean: [0.5, 0.5, 0.5]
+          std: [0.5, 0.5, 0.5]
+          order: hwc
+      - ToCHWImage: null
+      - KeepKeys:
+          keep_keys: [image, structure, bboxes, bbox_masks, shape]
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 10
+    num_workers: 8
\ No newline at end of file

configs/table/table_mv3.yml

@@ -4,7 +4,7 @@ Global:
   log_smooth_window: 20
   print_batch_step: 5
   save_model_dir: ./output/table_mv3/
-  save_epoch_step: 3
+  save_epoch_step: 400
   # evaluation is run every 400 iterations after the 0th iteration
   eval_batch_step: [0, 400]
   cal_metric_during_train: True
@@ -12,13 +12,12 @@ Global:
   checkpoints:
   save_inference_dir:
   use_visualdl: False
-  infer_img: doc/table/table.jpg
+  infer_img: ppstructure/docs/table/table.jpg
+  save_res_path: output/table_mv3
   # for data or label process
   character_dict_path: ppocr/utils/dict/table_structure_dict.txt
   character_type: en
-  max_text_length: 100
-  max_elem_length: 800
-  max_cell_num: 500
+  max_text_length: 800
   infer_mode: False
   process_total_num: 0
   process_cut_num: 0
@@ -44,11 +43,8 @@ Architecture:
   Head:
     name: TableAttentionHead
     hidden_size: 256
-    l2_decay: 0.00001
     loc_type: 2
-    max_text_length: 100
-    max_elem_length: 800
-    max_cell_num: 500
+    max_text_length: 800
 
 Loss:
   name: TableAttentionLoss
@@ -61,28 +57,34 @@ PostProcess:
 Metric:
   name: TableMetric
   main_indicator: acc
+  compute_bbox_metric: false # cost many time, set False for training
 
 Train:
   dataset:
     name: PubTabDataSet
     data_dir: train_data/table/pubtabnet/train/
-    label_file_path: train_data/table/pubtabnet/PubTabNet_2.0.0_train.jsonl
+    label_file_list: [train_data/table/pubtabnet/PubTabNet_2.0.0_train.jsonl]
     transforms:
       - DecodeImage: # load image
           img_mode: BGR
           channel_first: False
+      - TableLabelEncode:
+          learn_empty_box: False
+          merge_no_span_structure: False
+          replace_empty_cell_token: False
+      - TableBoxEncode:
       - ResizeTableImage:
           max_len: 488
-      - TableLabelEncode:
       - NormalizeImage:
           scale: 1./255.
           mean: [0.485, 0.456, 0.406]
           std: [0.229, 0.224, 0.225]
           order: 'hwc'
       - PaddingTableImage:
+          size: [488, 488]
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: ['image', 'structure', 'bbox_list', 'sp_tokens', 'bbox_list_mask']
+          keep_keys: [ 'image', 'structure', 'bboxes', 'bbox_masks', 'shape' ]
   loader:
     shuffle: True
     batch_size_per_card: 32
@@ -92,24 +94,29 @@ Train:
 Eval:
   dataset:
     name: PubTabDataSet
-    data_dir: train_data/table/pubtabnet/val/
-    label_file_path: train_data/table/pubtabnet/PubTabNet_2.0.0_val.jsonl
+    data_dir: /home/zhoujun20/table/PubTabNe/pubtabnet/val/
+    label_file_list: [/home/zhoujun20/table/PubTabNe/pubtabnet/val_500.jsonl]
     transforms:
       - DecodeImage: # load image
           img_mode: BGR
           channel_first: False
+      - TableLabelEncode:
+          learn_empty_box: False
+          merge_no_span_structure: False
+          replace_empty_cell_token: False
+      - TableBoxEncode:
       - ResizeTableImage:
           max_len: 488
-      - TableLabelEncode:
       - NormalizeImage:
           scale: 1./255.
           mean: [0.485, 0.456, 0.406]
           std: [0.229, 0.224, 0.225]
           order: 'hwc'
       - PaddingTableImage:
+          size: [488, 488]
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: ['image', 'structure', 'bbox_list', 'sp_tokens', 'bbox_list_mask']
+          keep_keys: [ 'image', 'structure', 'bboxes', 'bbox_masks', 'shape' ]
   loader:
     shuffle: False
     drop_last: False

configs/vqa/re/layoutlmv2_funsd.yml

+Global:
+  use_gpu: True
+  epoch_num: &epoch_num 200
+  log_smooth_window: 10
+  print_batch_step: 10
+  save_model_dir: ./output/re_layoutlmv2_funsd
+  save_epoch_step: 2000
+  # evaluation is run every 10 iterations after the 0th iteration
+  eval_batch_step: [ 0, 57 ]
+  cal_metric_during_train: False
+  save_inference_dir:
+  use_visualdl: False
+  seed: 2022
+  infer_img: train_data/FUNSD/testing_data/images/83624198.png
+  save_res_path: ./output/re_layoutlmv2_funsd/res/
+
+Architecture:
+  model_type: vqa
+  algorithm: &algorithm "LayoutLMv2"
+  Transform:
+  Backbone:
+    name: LayoutLMv2ForRe
+    pretrained: True
+    checkpoints:
+
+Loss:
+  name: LossFromOutput
+  key: loss
+  reduction: mean
+
+Optimizer:
+  name: AdamW
+  beta1: 0.9
+  beta2: 0.999
+  clip_norm: 10
+  lr:
+    learning_rate: 0.00005
+    warmup_epoch: 10
+  regularizer:
+    name: L2
+    factor: 0.00000
+    
+PostProcess:
+  name: VQAReTokenLayoutLMPostProcess
+
+Metric:
+  name: VQAReTokenMetric
+  main_indicator: hmean
+
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/FUNSD/training_data/images/
+    label_file_list: 
+      - ./train_data/FUNSD/train.json
+    ratio_list: [ 1.0 ]
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: True
+          algorithm: *algorithm
+          class_path: &class_path train_data/FUNSD/class_list.txt
+      - VQATokenPad:
+          max_seq_len: &max_seq_len 512
+          return_attention_mask: True
+      - VQAReTokenRelation:
+      - VQAReTokenChunk:
+          max_seq_len: *max_seq_len
+      - Resize:
+          size: [224,224]
+      - NormalizeImage:
+          scale: 1./255.
+          mean: [0.485, 0.456, 0.406]
+          std: [0.229, 0.224, 0.225]
+          order: 'hwc'
+      - ToCHWImage:
+      - KeepKeys:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'entities', 'relations']
+  loader:
+    shuffle: True
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 8
+    collate_fn: ListCollator
+
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/FUNSD/testing_data/images/
+    label_file_list: 
+      - ./train_data/FUNSD/test.json
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: True
+          algorithm: *algorithm
+          class_path: *class_path
+      - VQATokenPad:
+          max_seq_len: *max_seq_len
+          return_attention_mask: True
+      - VQAReTokenRelation:
+      - VQAReTokenChunk:
+          max_seq_len: *max_seq_len
+      - Resize:
+          size: [224,224]
+      - NormalizeImage:
+          scale: 1./255.
+          mean: [0.485, 0.456, 0.406]
+          std: [0.229, 0.224, 0.225]
+          order: 'hwc'
+      - ToCHWImage:
+      - KeepKeys:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'entities', 'relations']
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 8
+    collate_fn: ListCollator

configs/vqa/re/layoutlmv2.yml → configs/vqa/re/layoutlmv2_xund_zh.yml

@@ -3,16 +3,16 @@ Global:
   epoch_num: &epoch_num 200
   log_smooth_window: 10
   print_batch_step: 10
-  save_model_dir: ./output/re_layoutlmv2/
+  save_model_dir: ./output/re_layoutlmv2_xfund_zh
   save_epoch_step: 2000
   # evaluation is run every 10 iterations after the 0th iteration
-  eval_batch_step: [ 0, 19 ]
+  eval_batch_step: [ 0, 57 ]
   cal_metric_during_train: False
   save_inference_dir:
   use_visualdl: False
   seed: 2048
-  infer_img: doc/vqa/input/zh_val_21.jpg
-  save_res_path: ./output/re/
+  infer_img: ppstructure/docs/vqa/input/zh_val_21.jpg
+  save_res_path: ./output/re_layoutlmv2_xfund_zh/res/
 
 Architecture:
   model_type: vqa
@@ -52,7 +52,7 @@ Train:
     name: SimpleDataSet
     data_dir: train_data/XFUND/zh_train/image
     label_file_list: 
-      - train_data/XFUND/zh_train/xfun_normalize_train.json
+      - train_data/XFUND/zh_train/train.json
     ratio_list: [ 1.0 ]
     transforms:
       - DecodeImage: # load image
@@ -61,7 +61,7 @@ Train:
       - VQATokenLabelEncode: # Class handling label
           contains_re: True
           algorithm: *algorithm
-          class_path: &class_path ppstructure/vqa/labels/labels_ser.txt
+          class_path: &class_path train_data/XFUND/class_list_xfun.txt
       - VQATokenPad:
           max_seq_len: &max_seq_len 512
           return_attention_mask: True
@@ -77,7 +77,7 @@ Train:
           order: 'hwc'
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: [ 'input_ids', 'bbox', 'image', 'attention_mask', 'token_type_ids','entities', 'relations'] # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids','image', 'entities', 'relations'] # dataloader will return list in this order
   loader:
     shuffle: True
     drop_last: False
@@ -90,7 +90,7 @@ Eval:
     name: SimpleDataSet
     data_dir: train_data/XFUND/zh_val/image
     label_file_list:
-      - train_data/XFUND/zh_val/xfun_normalize_val.json
+      - train_data/XFUND/zh_val/val.json
     transforms:
       - DecodeImage: # load image
           img_mode: RGB
@@ -114,7 +114,7 @@ Eval:
           order: 'hwc'
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: [ 'input_ids', 'bbox', 'image', 'attention_mask', 'token_type_ids','entities', 'relations'] # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image','entities', 'relations'] # dataloader will return list in this order
   loader:
     shuffle: False
     drop_last: False

configs/vqa/re/layoutxlm_funsd.yml

+Global:
+  use_gpu: True
+  epoch_num: &epoch_num 200
+  log_smooth_window: 10
+  print_batch_step: 10
+  save_model_dir: ./output/re_layoutxlm_funsd
+  save_epoch_step: 2000
+  # evaluation is run every 10 iterations after the 0th iteration
+  eval_batch_step: [ 0, 57 ]
+  cal_metric_during_train: False
+  save_inference_dir:
+  use_visualdl: False
+  seed: 2022
+  infer_img: train_data/FUNSD/testing_data/images/83624198.png
+  save_res_path: ./output/re_layoutxlm_funsd/res/
+
+Architecture:
+  model_type: vqa
+  algorithm: &algorithm "LayoutXLM"
+  Transform:
+  Backbone:
+    name: LayoutXLMForRe
+    pretrained: True
+    checkpoints:
+
+Loss:
+  name: LossFromOutput
+  key: loss
+  reduction: mean
+
+Optimizer:
+  name: AdamW
+  beta1: 0.9
+  beta2: 0.999
+  clip_norm: 10
+  lr:
+    learning_rate: 0.00005
+    warmup_epoch: 10
+  regularizer:
+    name: L2
+    factor: 0.00000
+    
+PostProcess:
+  name: VQAReTokenLayoutLMPostProcess
+
+Metric:
+  name: VQAReTokenMetric
+  main_indicator: hmean
+
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/FUNSD/training_data/images/
+    label_file_list: 
+      - ./train_data/FUNSD/train_v4.json
+      # - ./train_data/FUNSD/train.json
+    ratio_list: [ 1.0 ]
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: True
+          algorithm: *algorithm
+          class_path: &class_path ./train_data/FUNSD/class_list.txt
+          use_textline_bbox_info: &use_textline_bbox_info True
+      - VQATokenPad:
+          max_seq_len: &max_seq_len 512
+          return_attention_mask: True
+      - VQAReTokenRelation:
+      - VQAReTokenChunk:
+          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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'entities', 'relations']
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 16
+    collate_fn: ListCollator
+
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/FUNSD/testing_data/images/
+    label_file_list: 
+      - ./train_data/FUNSD/test_v4.json
+      # - ./train_data/FUNSD/test.json
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: True
+          algorithm: *algorithm
+          class_path: *class_path
+          use_textline_bbox_info: *use_textline_bbox_info
+      - VQATokenPad:
+          max_seq_len: *max_seq_len
+          return_attention_mask: True
+      - VQAReTokenRelation:
+      - VQAReTokenChunk:
+          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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'entities', 'relations']
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 8
+    collate_fn: ListCollator

configs/vqa/re/layoutxlm.yml → configs/vqa/re/layoutxlm_xfund_zh.yml

@@ -11,7 +11,7 @@ Global:
   save_inference_dir:
   use_visualdl: False
   seed: 2022
-  infer_img: doc/vqa/input/zh_val_21.jpg
+  infer_img: ppstructure/docs/vqa/input/zh_val_21.jpg
   save_res_path: ./output/re/
 
 Architecture:
@@ -52,7 +52,7 @@ Train:
     name: SimpleDataSet
     data_dir: train_data/XFUND/zh_train/image
     label_file_list: 
-      - train_data/XFUND/zh_train/xfun_normalize_train.json
+      - train_data/XFUND/zh_train/train.json
     ratio_list: [ 1.0 ]
     transforms:
       - DecodeImage: # load image
@@ -61,7 +61,7 @@ Train:
       - VQATokenLabelEncode: # Class handling label
           contains_re: True
           algorithm: *algorithm
-          class_path: &class_path ppstructure/vqa/labels/labels_ser.txt
+          class_path: &class_path train_data/XFUND/class_list_xfun.txt
       - VQATokenPad:
           max_seq_len: &max_seq_len 512
           return_attention_mask: True
@@ -77,7 +77,7 @@ Train:
           order: 'hwc'
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: [ 'input_ids', 'bbox', 'image', 'attention_mask', 'token_type_ids','entities', 'relations'] # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox','attention_mask', 'token_type_ids', 'image', 'entities', 'relations'] # dataloader will return list in this order
   loader:
     shuffle: True
     drop_last: False
@@ -90,7 +90,7 @@ Eval:
     name: SimpleDataSet
     data_dir: train_data/XFUND/zh_val/image
     label_file_list:
-      - train_data/XFUND/zh_val/xfun_normalize_val.json
+      - train_data/XFUND/zh_val/val.json
     transforms:
       - DecodeImage: # load image
           img_mode: RGB
@@ -114,7 +114,7 @@ Eval:
           order: 'hwc'
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: [ 'input_ids', 'bbox', 'image', 'attention_mask', 'token_type_ids','entities', 'relations'] # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'entities', 'relations'] # dataloader will return list in this order
   loader:
     shuffle: False
     drop_last: False

configs/vqa/ser/layoutlm_funsd.yml

+Global:
+  use_gpu: True
+  epoch_num: &epoch_num 200
+  log_smooth_window: 10
+  print_batch_step: 10
+  save_model_dir: ./output/ser_layoutlm_funsd
+  save_epoch_step: 2000
+  # evaluation is run every 10 iterations after the 0th iteration
+  eval_batch_step: [ 0, 57 ]
+  cal_metric_during_train: False
+  save_inference_dir:
+  use_visualdl: False
+  seed: 2022
+  infer_img: train_data/FUNSD/testing_data/images/83624198.png
+  save_res_path: ./output/ser_layoutlm_funsd/res/
+
+Architecture:
+  model_type: vqa
+  algorithm: &algorithm "LayoutLM"
+  Transform:
+  Backbone:
+    name: LayoutLMForSer
+    pretrained: True
+    checkpoints:
+    num_classes: &num_classes 7
+
+Loss:
+  name: VQASerTokenLayoutLMLoss
+  num_classes: *num_classes
+
+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/FUNSD/class_list.txt
+
+Metric:
+  name: VQASerTokenMetric
+  main_indicator: hmean
+
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/FUNSD/training_data/images/
+    label_file_list: 
+      - ./train_data/FUNSD/train.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 True
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: True
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4
+
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: train_data/FUNSD/testing_data/images/
+    label_file_list:
+      - ./train_data/FUNSD/test.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
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4

configs/vqa/ser/layoutlm_sroie.yml

+Global:
+  use_gpu: True
+  epoch_num: &epoch_num 200
+  log_smooth_window: 10
+  print_batch_step: 10
+  save_model_dir: ./output/ser_layoutlm_sroie
+  save_epoch_step: 2000
+  # evaluation is run every 10 iterations after the 0th iteration
+  eval_batch_step: [ 0, 200 ]
+  cal_metric_during_train: False
+  save_inference_dir:
+  use_visualdl: False
+  seed: 2022
+  infer_img: train_data/SROIE/test/X00016469670.jpg
+  save_res_path: ./output/ser_layoutlm_sroie/res/
+
+Architecture:
+  model_type: vqa
+  algorithm: &algorithm "LayoutLM"
+  Transform:
+  Backbone:
+    name: LayoutLMForSer
+    pretrained: True
+    checkpoints:
+    num_classes: &num_classes 9
+
+Loss:
+  name: VQASerTokenLayoutLMLoss
+  num_classes: *num_classes
+
+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/SROIE/class_list.txt
+
+Metric:
+  name: VQASerTokenMetric
+  main_indicator: hmean
+
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/SROIE/train
+    label_file_list: 
+      - ./train_data/SROIE/train.txt
+    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
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: True
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4
+
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/SROIE/test
+    label_file_list: 
+      - ./train_data/SROIE/test.txt
+    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
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4

configs/vqa/ser/layoutlm.yml → configs/vqa/ser/layoutlm_xfund_zh.yml

@@ -3,16 +3,16 @@ Global:
   epoch_num: &epoch_num 200
   log_smooth_window: 10
   print_batch_step: 10
-  save_model_dir: ./output/ser_layoutlm/
+  save_model_dir: ./output/ser_layoutlm_xfund_zh
   save_epoch_step: 2000
   # evaluation is run every 10 iterations after the 0th iteration
-  eval_batch_step: [ 0, 19 ]
+  eval_batch_step: [ 0, 57 ]
   cal_metric_during_train: False
   save_inference_dir:
   use_visualdl: False
   seed: 2022
-  infer_img: doc/vqa/input/zh_val_0.jpg
-  save_res_path: ./output/ser/
+  infer_img: ppstructure/docs/vqa/input/zh_val_42.jpg
+  save_res_path: ./output/ser_layoutlm_xfund_zh/res/
 
 Architecture:
   model_type: vqa
@@ -43,7 +43,7 @@ Optimizer:
     
 PostProcess:
   name: VQASerTokenLayoutLMPostProcess
-  class_path: &class_path ppstructure/vqa/labels/labels_ser.txt
+  class_path: &class_path train_data/XFUND/class_list_xfun.txt
 
 Metric:
   name: VQASerTokenMetric
@@ -54,7 +54,7 @@ Train:
     name: SimpleDataSet
     data_dir: train_data/XFUND/zh_train/image
     label_file_list: 
-      - train_data/XFUND/zh_train/xfun_normalize_train.json
+      - train_data/XFUND/zh_train/train.json
     transforms:
       - DecodeImage: # load image
           img_mode: RGB
@@ -77,7 +77,7 @@ Train:
           order: 'hwc'
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: [ 'input_ids','labels', 'bbox', 'image', 'attention_mask', 'token_type_ids'] # dataloader will return list in this order
+          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
@@ -89,7 +89,7 @@ Eval:
     name: SimpleDataSet
     data_dir: train_data/XFUND/zh_val/image
     label_file_list:
-      - train_data/XFUND/zh_val/xfun_normalize_val.json
+      - train_data/XFUND/zh_val/val.json
     transforms:
       - DecodeImage: # load image
           img_mode: RGB
@@ -112,7 +112,7 @@ Eval:
           order: 'hwc'
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: [ 'input_ids', 'labels', 'bbox', 'image', 'attention_mask', 'token_type_ids'] # dataloader will return list in this order
+          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

configs/vqa/ser/layoutlmv2_funsd.yml

+Global:
+  use_gpu: True
+  epoch_num: &epoch_num 200
+  log_smooth_window: 10
+  print_batch_step: 10
+  save_model_dir: ./output/ser_layoutlmv2_funsd
+  save_epoch_step: 2000
+  # evaluation is run every 10 iterations after the 0th iteration
+  eval_batch_step: [ 0, 100 ]
+  cal_metric_during_train: False
+  save_inference_dir:
+  use_visualdl: False
+  seed: 2022
+  infer_img: train_data/FUNSD/testing_data/images/83624198.png
+  save_res_path: ./output/ser_layoutlmv2_funsd/res/
+
+Architecture:
+  model_type: vqa
+  algorithm: &algorithm "LayoutLMv2"
+  Transform:
+  Backbone:
+    name: LayoutLMv2ForSer
+    pretrained: True
+    checkpoints:
+    num_classes: &num_classes 7
+
+Loss:
+  name: VQASerTokenLayoutLMLoss
+  num_classes: *num_classes
+
+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/FUNSD/class_list.txt
+
+Metric:
+  name: VQASerTokenMetric
+  main_indicator: hmean
+
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/FUNSD/training_data/images/
+    label_file_list: 
+      - ./train_data/FUNSD/train.json
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: False
+          algorithm: *algorithm
+          class_path: *class_path
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: True
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4
+
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/FUNSD/testing_data/images/
+    label_file_list: 
+      - ./train_data/FUNSD/test.json
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: False
+          algorithm: *algorithm
+          class_path: *class_path
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4

configs/vqa/ser/layoutlmv2_sroie.yml

+Global:
+  use_gpu: True
+  epoch_num: &epoch_num 200
+  log_smooth_window: 10
+  print_batch_step: 10
+  save_model_dir: ./output/ser_layoutlmv2_sroie
+  save_epoch_step: 2000
+  # evaluation is run every 10 iterations after the 0th iteration
+  eval_batch_step: [ 0, 200 ]
+  cal_metric_during_train: False
+  save_inference_dir:
+  use_visualdl: False
+  seed: 2022
+  infer_img: train_data/SROIE/test/X00016469670.jpg
+  save_res_path: ./output/ser_layoutlmv2_sroie/res/
+
+Architecture:
+  model_type: vqa
+  algorithm: &algorithm "LayoutLMv2"
+  Transform:
+  Backbone:
+    name: LayoutLMv2ForSer
+    pretrained: True
+    checkpoints:
+    num_classes: &num_classes 9
+
+Loss:
+  name: VQASerTokenLayoutLMLoss
+  num_classes: *num_classes
+
+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/SROIE/class_list.txt
+
+Metric:
+  name: VQASerTokenMetric
+  main_indicator: hmean
+
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/SROIE/train
+    label_file_list: 
+      - ./train_data/SROIE/train.txt
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: False
+          algorithm: *algorithm
+          class_path: *class_path
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: True
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4
+
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/SROIE/test
+    label_file_list: 
+      - ./train_data/SROIE/test.txt
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: False
+          algorithm: *algorithm
+          class_path: *class_path
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4

configs/vqa/ser/layoutlmv2.yml → configs/vqa/ser/layoutlmv2_xfund_zh.yml

@@ -3,7 +3,7 @@ Global:
   epoch_num: &epoch_num 200
   log_smooth_window: 10
   print_batch_step: 10
-  save_model_dir: ./output/ser_layoutlmv2/
+  save_model_dir: ./output/ser_layoutlmv2_xfund_zh/
   save_epoch_step: 2000
   # evaluation is run every 10 iterations after the 0th iteration
   eval_batch_step: [ 0, 19 ]
@@ -11,8 +11,8 @@ Global:
   save_inference_dir:
   use_visualdl: False
   seed: 2022
-  infer_img: doc/vqa/input/zh_val_0.jpg
-  save_res_path: ./output/ser/
+  infer_img: ppstructure/docs/vqa/input/zh_val_42.jpg
+  save_res_path: ./output/ser_layoutlmv2_xfund_zh/res/
 
 Architecture:
   model_type: vqa
@@ -44,7 +44,7 @@ Optimizer:
     
 PostProcess:
   name: VQASerTokenLayoutLMPostProcess
-  class_path: &class_path ppstructure/vqa/labels/labels_ser.txt
+  class_path: &class_path train_data/XFUND/class_list_xfun.txt
 
 Metric:
   name: VQASerTokenMetric
@@ -55,7 +55,7 @@ Train:
     name: SimpleDataSet
     data_dir: train_data/XFUND/zh_train/image
     label_file_list: 
-      - train_data/XFUND/zh_train/xfun_normalize_train.json
+      - train_data/XFUND/zh_train/train.json
     transforms:
       - DecodeImage: # load image
           img_mode: RGB
@@ -78,7 +78,7 @@ Train:
           order: 'hwc'
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: [ 'input_ids','labels', 'bbox', 'image', 'attention_mask', 'token_type_ids'] # dataloader will return list in this order
+          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
@@ -90,7 +90,7 @@ Eval:
     name: SimpleDataSet
     data_dir: train_data/XFUND/zh_val/image
     label_file_list:
-      - train_data/XFUND/zh_val/xfun_normalize_val.json
+      - train_data/XFUND/zh_val/val.json
     transforms:
       - DecodeImage: # load image
           img_mode: RGB
@@ -113,7 +113,7 @@ Eval:
           order: 'hwc'
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: [ 'input_ids', 'labels', 'bbox', 'image', 'attention_mask', 'token_type_ids'] # dataloader will return list in this order
+          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

configs/vqa/ser/layoutxlm_funsd.yml

+Global:
+  use_gpu: True
+  epoch_num: &epoch_num 200
+  log_smooth_window: 10
+  print_batch_step: 10
+  save_model_dir: ./output/ser_layoutxlm_funsd
+  save_epoch_step: 2000
+  # evaluation is run every 10 iterations after the 0th iteration
+  eval_batch_step: [ 0, 57 ]
+  cal_metric_during_train: False
+  save_inference_dir:
+  use_visualdl: False
+  seed: 2022
+  infer_img: train_data/FUNSD/testing_data/images/83624198.png
+  save_res_path: output/ser_layoutxlm_funsd/res/
+
+Architecture:
+  model_type: vqa
+  algorithm: &algorithm "LayoutXLM"
+  Transform:
+  Backbone:
+    name: LayoutXLMForSer
+    pretrained: True
+    checkpoints:
+    num_classes: &num_classes 7
+
+Loss:
+  name: VQASerTokenLayoutLMLoss
+  num_classes: *num_classes
+
+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/FUNSD/class_list.txt
+
+Metric:
+  name: VQASerTokenMetric
+  main_indicator: hmean
+
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/FUNSD/training_data/images/
+    label_file_list: 
+      - ./train_data/FUNSD/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
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: True
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4
+
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: train_data/FUNSD/testing_data/images/
+    label_file_list:
+      - ./train_data/FUNSD/test.json
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: False
+          algorithm: *algorithm
+          class_path: *class_path
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4

configs/vqa/ser/layoutxlm_sroie.yml

+Global:
+  use_gpu: True
+  epoch_num: &epoch_num 200
+  log_smooth_window: 10
+  print_batch_step: 10
+  save_model_dir: ./output/ser_layoutxlm_sroie
+  save_epoch_step: 2000
+  # evaluation is run every 10 iterations after the 0th iteration
+  eval_batch_step: [ 0, 200 ]
+  cal_metric_during_train: False
+  save_inference_dir:
+  use_visualdl: False
+  seed: 2022
+  infer_img: train_data/SROIE/test/X00016469670.jpg
+  save_res_path: res_img_aug_with_gt
+
+Architecture:
+  model_type: vqa
+  algorithm: &algorithm "LayoutXLM"
+  Transform:
+  Backbone:
+    name: LayoutXLMForSer
+    pretrained: True
+    checkpoints:
+    num_classes: &num_classes 9
+
+Loss:
+  name: VQASerTokenLayoutLMLoss
+  num_classes: *num_classes
+
+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/SROIE/class_list.txt
+
+Metric:
+  name: VQASerTokenMetric
+  main_indicator: hmean
+
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/SROIE/train
+    label_file_list: 
+      - ./train_data/SROIE/train.txt
+    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
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: True
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4
+
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: train_data/SROIE/test
+    label_file_list:
+      - ./train_data/SROIE/test.txt
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: False
+          algorithm: *algorithm
+          class_path: *class_path
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4

configs/vqa/ser/layoutxlm_wildreceipt.yml

+Global:
+  use_gpu: True
+  epoch_num: &epoch_num 100
+  log_smooth_window: 10
+  print_batch_step: 10
+  save_model_dir: ./output/ser_layoutxlm_wildreceipt
+  save_epoch_step: 2000
+  # evaluation is run every 10 iterations after the 0th iteration
+  eval_batch_step: [ 0, 200 ]
+  cal_metric_during_train: False
+  save_inference_dir:
+  use_visualdl: False
+  seed: 2022
+  infer_img: train_data//wildreceipt/image_files/Image_12/10/845be0dd6f5b04866a2042abd28d558032ef2576.jpeg
+  save_res_path: ./output/ser_layoutxlm_wildreceipt/res
+
+Architecture:
+  model_type: vqa
+  algorithm: &algorithm "LayoutXLM"
+  Transform:
+  Backbone:
+    name: LayoutXLMForSer
+    pretrained: True
+    checkpoints:
+    num_classes: &num_classes 51
+
+Loss:
+  name: VQASerTokenLayoutLMLoss
+  num_classes: *num_classes
+
+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/wildreceipt/class_list.txt
+
+Metric:
+  name: VQASerTokenMetric
+  main_indicator: hmean
+
+Train:
+  dataset:
+    name: SimpleDataSet
+    data_dir: ./train_data/wildreceipt/
+    label_file_list: 
+      - ./train_data/wildreceipt/wildreceipt_train.txt
+    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
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: True
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4
+
+Eval:
+  dataset:
+    name: SimpleDataSet
+    data_dir: train_data/wildreceipt
+    label_file_list:
+      - ./train_data/wildreceipt/wildreceipt_test.txt
+    transforms:
+      - DecodeImage: # load image
+          img_mode: RGB
+          channel_first: False
+      - VQATokenLabelEncode: # Class handling label
+          contains_re: False
+          algorithm: *algorithm
+          class_path: *class_path
+      - 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:
+          # dataloader will return list in this order
+          keep_keys: [ 'input_ids', 'bbox', 'attention_mask', 'token_type_ids', 'image', 'labels']
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 8
+    num_workers: 4

configs/vqa/ser/layoutxlm.yml → configs/vqa/ser/layoutxlm_xfund_zh.yml

@@ -3,7 +3,7 @@ Global:
   epoch_num: &epoch_num 200
   log_smooth_window: 10
   print_batch_step: 10
-  save_model_dir: ./output/ser_layoutxlm/
+  save_model_dir: ./output/ser_layoutxlm_xfund_zh
   save_epoch_step: 2000
   # evaluation is run every 10 iterations after the 0th iteration
   eval_batch_step: [ 0, 19 ]
@@ -11,8 +11,8 @@ Global:
   save_inference_dir:
   use_visualdl: False
   seed: 2022
-  infer_img: doc/vqa/input/zh_val_42.jpg
-  save_res_path: ./output/ser
+  infer_img: ppstructure/docs/vqa/input/zh_val_42.jpg
+  save_res_path: ./output/ser_layoutxlm_xfund_zh/res
 
 Architecture:
   model_type: vqa
@@ -43,7 +43,7 @@ Optimizer:
     
 PostProcess:
   name: VQASerTokenLayoutLMPostProcess
-  class_path: &class_path ppstructure/vqa/labels/labels_ser.txt
+  class_path: &class_path train_data/XFUND/class_list_xfun.txt
 
 Metric:
   name: VQASerTokenMetric
@@ -54,7 +54,7 @@ Train:
     name: SimpleDataSet
     data_dir: train_data/XFUND/zh_train/image
     label_file_list: 
-      - train_data/XFUND/zh_train/xfun_normalize_train.json
+      - train_data/XFUND/zh_train/train.json
     ratio_list: [ 1.0 ]
     transforms:
       - DecodeImage: # load image
@@ -78,7 +78,7 @@ Train:
           order: 'hwc'
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: [ 'input_ids','labels', 'bbox', 'image', 'attention_mask', 'token_type_ids'] # dataloader will return list in this order
+          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
@@ -90,7 +90,7 @@ Eval:
     name: SimpleDataSet
     data_dir: train_data/XFUND/zh_val/image
     label_file_list:
-      - train_data/XFUND/zh_val/xfun_normalize_val.json
+      - train_data/XFUND/zh_val/val.json
     transforms:
       - DecodeImage: # load image
           img_mode: RGB
@@ -113,7 +113,7 @@ Eval:
           order: 'hwc'
       - ToCHWImage:
       - KeepKeys:
-          keep_keys: [ 'input_ids', 'labels', 'bbox', 'image', 'attention_mask', 'token_type_ids'] # dataloader will return list in this order
+          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

deploy/avh/README.md

@@ -15,20 +15,24 @@
 <!--- specific language governing permissions and limitations -->
 <!--- under the License. -->
 
-Running PaddleOCR text recognition model via TVM on bare metal Arm(R) Cortex(R)-M55 CPU and CMSIS-NN
-===============================================================
+English | [简体中文](README_ch.md)
 
-This folder contains an example of how to use TVM to run a PaddleOCR model
-on bare metal Cortex(R)-M55 CPU and CMSIS-NN.
+Running PaddleOCR text recognition model on bare metal Arm(R) Cortex(R)-M55 CPU using Arm Virtual Hardware
+======================================================================
 
-Prerequisites
+This folder contains an example of how to run a PaddleOCR model on bare metal [Cortex(R)-M55 CPU](https://www.arm.com/products/silicon-ip-cpu/cortex-m/cortex-m55) using [Arm Virtual Hardware](https://www.arm.com/products/development-tools/simulation/virtual-hardware).
+
+
+Running environment and prerequisites
 -------------
-If the demo is run in the ci_cpu Docker container provided with TVM, then the following
-software will already be installed.
+Case 1: If the demo is run in Arm Virtual Hardware Amazon Machine Image(AMI) instance hosted by [AWS](https://aws.amazon.com/marketplace/pp/prodview-urbpq7yo5va7g?sr=0-1&ref_=beagle&applicationId=AWSMPContessa)/[AWS China](https://awsmarketplace.amazonaws.cn/marketplace/pp/prodview-2y7nefntbmybu), the following software will be installed through [configure_avh.sh](./configure_avh.sh) script. It will install automatically when you run the application through [run_demo.sh](./run_demo.sh) script.
+You can refer to this [guide](https://arm-software.github.io/AVH/main/examples/html/MicroSpeech.html#amilaunch) to launch an Arm Virtual Hardware AMI instance.
+
+Case 2: If the demo is run in the [ci_cpu Docker container](https://github.com/apache/tvm/blob/main/docker/Dockerfile.ci_cpu) provided with [TVM](https://github.com/apache/tvm), then the following software will already be installed.
 
-If the demo is not run in the ci_cpu Docker container, then you will need the following:
+Case 3: If the demo is not run in the ci_cpu Docker container, then you will need the following:
 - Software required to build and run the demo (These can all be installed by running 
-  https://github.com/apache/tvm/blob/main/docker/install/ubuntu_install_ethosu_driver_stack.sh .)
+  tvm/docker/install/ubuntu_install_ethosu_driver_stack.sh.)
   - [Fixed Virtual Platform (FVP) based on Arm(R) Corstone(TM)-300 software](https://developer.arm.com/tools-and-software/open-source-software/arm-platforms-software/arm-ecosystem-fvps)
   - [cmake 3.19.5](https://github.com/Kitware/CMake/releases/)
   - [GCC toolchain from Arm(R)](https://developer.arm.com/-/media/Files/downloads/gnu-rm/10-2020q4/gcc-arm-none-eabi-10-2020-q4-major-x86_64-linux.tar.bz2)
@@ -40,19 +44,22 @@ If the demo is not run in the ci_cpu Docker container, then you will need the fo
     pip install -r ./requirements.txt
     ```
 
+In case2 and case3:  
+    
+You will need to update your PATH environment variable to include the path to cmake 3.19.5 and the FVP.
+For example if you've installed these in ```/opt/arm``` , then you would do the following:
+```bash
+export PATH=/opt/arm/FVP_Corstone_SSE-300/models/Linux64_GCC-6.4:/opt/arm/cmake/bin:$PATH
+```
+
 You will also need TVM which can either be:
+  - Installed from TLCPack(see [TLCPack](https://tlcpack.ai/))
   - Built from source (see [Install from Source](https://tvm.apache.org/docs/install/from_source.html))
     - When building from source, the following need to be set in config.cmake:
       - set(USE_CMSISNN ON)
       - set(USE_MICRO ON)
       - set(USE_LLVM ON)
-  - Installed from TLCPack nightly(see [TLCPack](https://tlcpack.ai/))
 
-You will need to update your PATH environment variable to include the path to cmake 3.19.5 and the FVP.
-For example if you've installed these in ```/opt/arm``` , then you would do the following:
-```bash
-export PATH=/opt/arm/FVP_Corstone_SSE-300/models/Linux64_GCC-6.4:/opt/arm/cmake/bin:$PATH
-```
 
 Running the demo application
 ----------------------------
@@ -62,6 +69,12 @@ Type the following command to run the bare metal text recognition application ([
 ./run_demo.sh
 ```
 
+If you are not able to use Arm Virtual Hardware Amazon Machine Image(AMI) instance hosted by AWS/AWS China, specify argument --enable_FVP to 1 to make the application run on local Fixed Virtual Platforms (FVPs) executables.
+
+```bash
+./run_demo.sh --enable_FVP 1
+```
+
 If the Ethos(TM)-U platform and/or CMSIS have not been installed in /opt/arm/ethosu then
 the locations for these can be specified as arguments to run_demo.sh, for example:
 
@@ -70,13 +83,14 @@ the locations for these can be specified as arguments to run_demo.sh, for exampl
 --ethosu_platform_path /home/tvm-user/ethosu/core_platform
 ```
 
-This will:
+With [run_demo.sh](./run_demo.sh) to run the demo application, it will:
+- Set up running environment by installing the required prerequisites automatically if running in Arm Virtual Hardware Amazon AMI instance(not specify --enable_FVP to 1)
 - Download a PaddleOCR text recognition model
 - Use tvmc to compile the text recognition model for Cortex(R)-M55 CPU and CMSIS-NN
 - Create a C header file inputs.c containing the image data as a C array
 - Create a C header file outputs.c containing a C array where the output of inference will be stored
 - Build the demo application
-- Run the demo application on a Fixed Virtual Platform (FVP) based on Arm(R) Corstone(TM)-300 software
+- Run the demo application on a Arm Virtual Hardware based on Arm(R) Corstone(TM)-300 software
 - The application will report the text on the image and the corresponding score.
 
 Using your own image
@@ -92,9 +106,11 @@ python3 ./convert_image.py path/to/image
 
 Model description
 -----------------
-In this demo, the model we use is an English recognition model based on [PP-OCRv3](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/PP-OCRv3_introduction.md). PP-OCRv3 is the third version of the PP-OCR series model released by [PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR). This series of models has the following features:
+The example is built on [PP-OCRv3](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/PP-OCRv3_introduction.md) English recognition model released by [PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR). Since Arm(R) Cortex(R)-M55 CPU does not support rnn operator, we delete the unsupported operator based on the PP-OCRv3 text recognition model to obtain the current 2.7M English recognition model.
+
+PP-OCRv3 is the third version of the PP-OCR series model. This series of models has the following features:
   - PP-OCRv3: ultra-lightweight OCR system: detection (3.6M) + direction classifier (1.4M) + recognition (12M) = 17.0M
   - Support more than 80 kinds of multi-language recognition models, including English, Chinese, French, German, Arabic, Korean, Japanese and so on. For details
   - Support vertical text recognition, and long text recognition
 
-The text recognition model in PP-OCRv3 supports more than 80 languages. In the process of model development, since Arm(R) Cortex(R)-M55 CPU does not support rnn operator, we delete the unsupported operator based on the PP-OCRv3 text recognition model to obtain the current model.
\ No newline at end of file
+

deploy/avh/README_ch.md

@@ -14,6 +14,7 @@
 <!--- KIND, either express or implied.  See the License for the -->
 <!--- specific language governing permissions and limitations -->
 <!--- under the License. -->
+[English](README.md) | 简体中文
 
 通过TVM在 Arm(R) Cortex(R)-M55 CPU 上运行 PaddleOCR文 本能识别模型
 ===============================================================
@@ -85,9 +86,9 @@ export PATH=/opt/arm/FVP_Corstone_SSE-300/models/Linux64_GCC-6.4:/opt/arm/cmake/
 模型描述
 -----------------
 
-在这个demo中，我们使用的模型是基于[PP-OCRv3](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/PP-OCRv3_introduction.md)的英文识别模型。 PP-OCRv3是[PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR)发布的PP-OCR系列模型的第三个版本。 该系列模型具有以下特点：
+在这个demo中，我们使用的模型是基于[PP-OCRv3](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/PP-OCRv3_introduction.md)的英文识别模型。由于Arm(R) Cortex(R)-M55 CPU不支持rnn算子，我们在PP-OCRv3原始文本识别模型的基础上进行适配，最终模型大小为2.7M。
+
+PP-OCRv3是[PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR)发布的PP-OCR系列模型的第三个版本，该系列模型具有以下特点：
    - 超轻量级OCR系统：检测（3.6M）+方向分类器（1.4M）+识别（12M）=17.0M。
    - 支持80多种多语言识别模型，包括英文、中文、法文、德文、阿拉伯文、韩文、日文等。 
    - 支持竖排文本识别，长文本识别。
-
-PP-OCRv3 中的文本识别模型支持 80 多种语言。 在模型开发过程中，由于Arm(R) Cortex(R)-M55 CPU不支持rnn算子，我们在PP-OCRv3文本识别模型的基础上删除了不支持的算子，得到当前模型。
\ No newline at end of file

deploy/avh/configure_avh.sh

+#!/bin/bash
+# Copyright (c) 2022 Arm Limited and Contributors. All rights reserved.
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you 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.
+
+set -e
+set -u
+set -o pipefail
+
+# Show usage
+function show_usage() {
+    cat <<EOF
+Usage: Set up running environment by installing the required prerequisites.
+-h, --help
+    Display this help message.
+EOF
+}
+
+if [ "$#" -eq 1 ] && [ "$1" == "--help" -o "$1" == "-h" ]; then
+    show_usage
+    exit 0
+elif [ "$#" -ge 1 ]; then
+        show_usage
+        exit 1
+fi
+
+echo -e "\e[36mStart setting up running environment\e[0m"
+
+# Install CMSIS
+echo -e "\e[36mStart installing CMSIS\e[0m"
+CMSIS_PATH="/opt/arm/ethosu/cmsis" 
+mkdir -p "${CMSIS_PATH}"
+
+CMSIS_SHA="977abe9849781a2e788b02282986480ff4e25ea6"
+CMSIS_SHASUM="86c88d9341439fbb78664f11f3f25bc9fda3cd7de89359324019a4d87d169939eea85b7fdbfa6ad03aa428c6b515ef2f8cd52299ce1959a5444d4ac305f934cc"
+CMSIS_URL="http://github.com/ARM-software/CMSIS_5/archive/${CMSIS_SHA}.tar.gz"
+DOWNLOAD_PATH="/tmp/${CMSIS_SHA}.tar.gz"
+
+wget ${CMSIS_URL} -O "${DOWNLOAD_PATH}"
+echo "$CMSIS_SHASUM" ${DOWNLOAD_PATH} | sha512sum -c
+tar -xf "${DOWNLOAD_PATH}" -C "${CMSIS_PATH}" --strip-components=1
+touch "${CMSIS_PATH}"/"${CMSIS_SHA}".sha
+echo -e "\e[36mCMSIS Installation SUCCESS\e[0m"
+
+# Install Arm(R) Ethos(TM)-U NPU driver stack
+echo -e "\e[36mStart installing Arm(R) Ethos(TM)-U NPU driver stack\e[0m"
+git clone "https://review.mlplatform.org/ml/ethos-u/ethos-u-core-platform" /opt/arm/ethosu/core_platform 
+cd /opt/arm/ethosu/core_platform 
+git checkout tags/"21.11"
+echo -e "\e[36mArm(R) Ethos(TM)-U Core Platform Installation SUCCESS\e[0m"
+ 
+# Install Arm(R) GNU Toolchain
+echo -e "\e[36mStart installing Arm(R) GNU Toolchain\e[0m"
+mkdir -p /opt/arm/gcc-arm-none-eabi
+export gcc_arm_url='https://developer.arm.com/-/media/Files/downloads/gnu-rm/10-2020q4/gcc-arm-none-eabi-10-2020-q4-major-x86_64-linux.tar.bz2?revision=ca0cbf9c-9de2-491c-ac48-898b5bbc0443&la=en&hash=68760A8AE66026BCF99F05AC017A6A50C6FD832A'
+curl --retry 64 -sSL ${gcc_arm_url} | tar -C /opt/arm/gcc-arm-none-eabi --strip-components=1 -jx 
+export PATH=/opt/arm/gcc-arm-none-eabi/bin:$PATH 
+arm-none-eabi-gcc --version
+arm-none-eabi-g++ --version
+echo -e "\e[36mArm(R) Arm(R) GNU Toolchain Installation SUCCESS\e[0m"
+
+# Install TVM from TLCPack
+echo -e "\e[36mStart installing TVM\e[0m"
+pip install tlcpack-nightly -f https://tlcpack.ai/wheels
+echo -e "\e[36mTVM Installation SUCCESS\e[0m"
\ No newline at end of file

deploy/avh/run_demo.sh

@@ -33,9 +33,19 @@ Usage: run_demo.sh
 	Set path to FVP.
 --cmake_path
 	Set path to cmake.
+--enable_FVP
+	Set 1 to run application on local Fixed Virtual Platforms (FVPs) executables.
 EOF
 }
 
+# Configure environment variables
+FVP_enable=0
+export PATH=/opt/arm/gcc-arm-none-eabi/bin:$PATH
+
+# Install python libraries
+echo -e "\e[36mInstall python libraries\e[0m"
+sudo pip install -r ./requirements.txt
+
 # Parse arguments
 while (( $# )); do
     case "$1" in
@@ -92,6 +102,18 @@ while (( $# )); do
             fi
             ;;
             
+        --enable_FVP)
+            if [ $# -gt 1 ] && [ "$2" == "1" -o "$2" == "0" ];
+            then
+                FVP_enable="$2"
+                shift 2
+            else
+                echo 'ERROR: --enable_FVP requires a right argument 1 or 0' >&2
+                show_usage >&2
+                exit 1
+            fi
+            ;;
+
         -*|--*)
             echo "Error: Unknown flag: $1" >&2
             show_usage >&2
@@ -100,17 +122,27 @@ while (( $# )); do
     esac
 done
 
+# Choose running environment: cloud(default) or local environment
+Platform="VHT_Corstone_SSE-300_Ethos-U55"
+if [ $FVP_enable == "1" ]; then
+	Platform="FVP_Corstone_SSE-300_Ethos-U55"
+	echo -e "\e[36mRun application on local Fixed Virtual Platforms (FVPs)\e[0m"
+else
+	if [ ! -d "/opt/arm/" ]; then
+		sudo ./configure_avh.sh
+	fi
+fi
 
 # Directories
 script_dir="$( cd "$( dirname "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )"
 
-
 # Make build directory
-rm -rf build
 make cleanall
 mkdir -p build
 cd build
 
+# Get PaddlePaddle inference model
+echo -e "\e[36mDownload PaddlePaddle inference model\e[0m"
 wget https://paddleocr.bj.bcebos.com/tvm/ocr_en.tar
 tar -xf ocr_en.tar
 
@@ -144,9 +176,9 @@ cd ${script_dir}
 echo ${script_dir}
 make
 
-# Run demo executable on the FVP
-FVP_Corstone_SSE-300_Ethos-U55 -C cpu0.CFGDTCMSZ=15 \
+# Run demo executable on the AVH
+$Platform -C cpu0.CFGDTCMSZ=15 \
 -C cpu0.CFGITCMSZ=15 -C mps3_board.uart0.out_file=\"-\" -C mps3_board.uart0.shutdown_tag=\"EXITTHESIM\" \
 -C mps3_board.visualisation.disable-visualisation=1 -C mps3_board.telnetterminal0.start_telnet=0 \
 -C mps3_board.telnetterminal1.start_telnet=0 -C mps3_board.telnetterminal2.start_telnet=0 -C mps3_board.telnetterminal5.start_telnet=0 \
-./build/demo
\ No newline at end of file
+./build/demo --stat

deploy/lite/crnn_process.cc

@@ -34,12 +34,13 @@ cv::Mat CrnnResizeImg(cv::Mat img, float wh_ratio, int rec_image_height) {
     resize_w = imgW;
   else
     resize_w = int(ceilf(imgH * ratio));
-
+  cv::Mat resize_img;
   cv::resize(img, resize_img, cv::Size(resize_w, imgH), 0.f, 0.f,
              cv::INTER_LINEAR);
   cv::copyMakeBorder(resize_img, resize_img, 0, 0, 0,
                      int(imgW - resize_img.cols), cv::BORDER_CONSTANT,
                      {127, 127, 127});
+  return resize_img;
 }
 
 std::vector<std::string> ReadDict(std::string path) {

deploy/lite/ocr_db_crnn.cc

@@ -474,7 +474,7 @@ void system(char **argv){
   
     std::vector<double> rec_times;
     RunRecModel(boxes, srcimg, rec_predictor, rec_text, rec_text_score,
-                charactor_dict, cls_predictor, use_direction_classify, &rec_times);
+                charactor_dict, cls_predictor, use_direction_classify, &rec_times, rec_image_height);
   
     //// visualization
     auto img_vis = Visualization(srcimg, boxes);

doc/doc_ch/algorithm.md

@@ -5,6 +5,7 @@ PaddleOCR将**持续新增**支持OCR领域前沿算法与模型，已支持的
 - [文本检测算法](./algorithm_overview.md#11-%E6%96%87%E6%9C%AC%E6%A3%80%E6%B5%8B%E7%AE%97%E6%B3%95)
 - [文本识别算法](./algorithm_overview.md#12-%E6%96%87%E6%9C%AC%E8%AF%86%E5%88%AB%E7%AE%97%E6%B3%95)
 - [端到端算法](./algorithm_overview.md#2-%E6%96%87%E6%9C%AC%E8%AF%86%E5%88%AB%E7%AE%97%E6%B3%95)
+- [表格识别]](./algorithm_overview.md#3-%E8%A1%A8%E6%A0%BC%E8%AF%86%E5%88%AB%E7%AE%97%E6%B3%95)
 
 **欢迎广大开发者合作共建，贡献更多算法，合入有奖🎁！具体可查看[社区常规赛](https://github.com/PaddlePaddle/PaddleOCR/issues/4982)。**
 

doc/doc_ch/algorithm_det_db.md

-# DB
+# DB与DB++
 
 - [1. 算法简介](#1)
 - [2. 环境配置](#2)
@@ -21,12 +21,24 @@
 > Liao, Minghui and Wan, Zhaoyi and Yao, Cong and Chen, Kai and Bai, Xiang
 > AAAI, 2020
 
+> [Real-Time Scene Text Detection with Differentiable Binarization and Adaptive Scale Fusion](https://arxiv.org/abs/2202.10304)
+> Liao, Minghui and Zou, Zhisheng and Wan, Zhaoyi and Yao, Cong and Bai, Xiang
+> TPAMI, 2022
+
+
 在ICDAR2015文本检测公开数据集上，算法复现效果如下：
 
 |模型|骨干网络|配置文件|precision|recall|Hmean|下载链接|
 | --- | --- | --- | --- | --- | --- | --- |
 |DB|ResNet50_vd|[configs/det/det_r50_vd_db.yml](../../configs/det/det_r50_vd_db.yml)|86.41%|78.72%|82.38%|[训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_db_v2.0_train.tar)|
 |DB|MobileNetV3|[configs/det/det_mv3_db.yml](../../configs/det/det_mv3_db.yml)|77.29%|73.08%|75.12%|[训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_mv3_db_v2.0_train.tar)|
+|DB++|ResNet50|[configs/det/det_r50_db++_icdar15.yml](../../configs/det/det_r50_db++_icdar15.yml)|90.89%|82.66%|86.58%|[合成数据预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.1/en_det/ResNet50_dcn_asf_synthtext_pretrained.pdparams)/[训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.1/en_det/det_r50_db%2B%2B_icdar15_train.tar)|
+
+在TD_TR文本检测公开数据集上，算法复现效果如下：
+
+|模型|骨干网络|配置文件|precision|recall|Hmean|下载链接|
+| --- | --- | --- | --- | --- | --- | --- |
+|DB++|ResNet50|[configs/det/det_r50_db++_td_tr.yml](../../configs/det/det_r50_db++_td_tr.yml)|92.92%|86.48%|89.58%|[合成数据预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.1/en_det/ResNet50_dcn_asf_synthtext_pretrained.pdparams)/[训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.1/en_det/det_r50_db%2B%2B_td_tr_train.tar)|
 
 
 <a name="2"></a>
@@ -54,7 +66,7 @@ python3 tools/export_model.py -c configs/det/det_r50_vd_db.yml -o Global.pretrai
 DB文本检测模型推理，可以执行如下命令：
 
 ```shell
-python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_model_dir="./inference/det_db/"
+python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_model_dir="./inference/det_db/" --det_algorithm="DB"
 ```
 
 可视化文本检测结果默认保存到`./inference_results`文件夹里面，结果文件的名称前缀为'det_res'。结果示例如下：
@@ -96,4 +108,12 @@ DB模型还支持以下推理部署方式：
   pages={11474--11481},
   year={2020}
 }
+
+@article{liao2022real,
+  title={Real-Time Scene Text Detection with Differentiable Binarization and Adaptive Scale Fusion},
+  author={Liao, Minghui and Zou, Zhisheng and Wan, Zhaoyi and Yao, Cong and Bai, Xiang},
+  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
+  year={2022},
+  publisher={IEEE}
+}
 ```

doc/doc_ch/algorithm_det_fcenet.md

 # FCENet
 
-- [1. 算法简介](#1)
-- [2. 环境配置](#2)
-- [3. 模型训练、评估、预测](#3)
-    - [3.1 训练](#3-1)
-    - [3.2 评估](#3-2)
-    - [3.3 预测](#3-3)
-- [4. 推理部署](#4)
-    - [4.1 Python推理](#4-1)
-    - [4.2 C++推理](#4-2)
-    - [4.3 Serving服务化部署](#4-3)
-    - [4.4 更多推理部署](#4-4)
-- [5. FAQ](#5)
+- [1. 算法简介](#1-算法简介)
+- [2. 环境配置](#2-环境配置)
+- [3. 模型训练、评估、预测](#3-模型训练评估预测)
+- [4. 推理部署](#4-推理部署)
+  - [4.1 Python推理](#41-python推理)
+  - [4.2 C++推理](#42-c推理)
+  - [4.3 Serving服务化部署](#43-serving服务化部署)
+  - [4.4 更多推理部署](#44-更多推理部署)
+- [5. FAQ](#5-faq)
+- [引用](#引用)
 
 <a name="1"></a>
 ## 1. 算法简介

doc/doc_ch/algorithm_overview.md

 # OCR算法
 
-- [1. 两阶段算法](#1-两阶段算法)
-    - [1.1 文本检测算法](#11-文本检测算法)
-    - [1.2 文本识别算法](#12-文本识别算法)
-- [2. 端到端算法](#2-端到端算法)
+- [1. 两阶段算法](#1)
+  - [1.1 文本检测算法](#11)
+  - [1.2 文本识别算法](#12)
+- [2. 端到端算法](#2)
+- [3. 表格识别算法](#3)
 
 
 本文给出了PaddleOCR已支持的OCR算法列表，以及每个算法在**英文公开数据集**上的模型和指标，主要用于算法简介和算法性能对比，更多包括中文在内的其他数据集上的模型请参考[PP-OCR v2.0 系列模型下载](./models_list.md)。
@@ -86,8 +87,9 @@
 |SAR|Resnet31| 87.20% | rec_r31_sar | [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.1/rec/rec_r31_sar_train.tar) |
 |SEED|Aster_Resnet| 85.35% | rec_resnet_stn_bilstm_att | [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.1/rec/rec_resnet_stn_bilstm_att.tar) |
 |SVTR|SVTR-Tiny| 89.25% | rec_svtr_tiny_none_ctc_en | [训练模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/rec_svtr_tiny_none_ctc_en_train.tar) |
-|ViTSTR|ViTSTR| 79.82% | rec_vitstr_none_ce_en | [训练模型](https://paddleocr.bj.bcebos.com/rec_vitstr_none_ce_train.tar) |
-|ABINet|Resnet45| 90.75% | rec_r45_abinet_en | [训练模型](https://paddleocr.bj.bcebos.com/rec_r45_abinet_train.tar) |
+|ViTSTR|ViTSTR| 79.82% | rec_vitstr_none_ce | [训练模型](https://paddleocr.bj.bcebos.com/rec_vitstr_none_ce_train.tar) |
+|ABINet|Resnet45| 90.75% | rec_r45_abinet | [训练模型](https://paddleocr.bj.bcebos.com/rec_r45_abinet_train.tar) |
+
 
 <a name="2"></a>
 
@@ -95,3 +97,16 @@
 
 已支持的端到端OCR算法列表（戳链接获取使用教程）：
 - [x]  [PGNet](./algorithm_e2e_pgnet.md)
+
+<a name="3"></a>
+
+## 3. 表格识别算法
+
+已支持的表格识别算法列表（戳链接获取使用教程）：
+- [x]  [TableMaster](./algorithm_table_master.md)
+
+在PubTabNet表格识别公开数据集上，算法效果如下：
+
+|模型|骨干网络|配置文件|acc|下载链接|
+|---|---|---|---|---|
+|TableMaster|TableResNetExtra|[configs/table/table_master.yml](../../configs/table/table_master.yml)|77.47%|[训练模型](https://paddleocr.bj.bcebos.com/ppstructure/models/tablemaster/table_structure_tablemaster_train.tar) / [推理模型](https://paddleocr.bj.bcebos.com/ppstructure/models/tablemaster/table_structure_tablemaster_infer.tar)|

doc/doc_ch/algorithm_table_master.md

+# 表格识别算法-TableMASTER
+
+- [1. 算法简介](#1-算法简介)
+- [2. 环境配置](#2-环境配置)
+- [3. 模型训练、评估、预测](#3-模型训练评估预测)
+- [4. 推理部署](#4-推理部署)
+  - [4.1 Python推理](#41-python推理)
+  - [4.2 C++推理部署](#42-c推理部署)
+  - [4.3 Serving服务化部署](#43-serving服务化部署)
+  - [4.4 更多推理部署](#44-更多推理部署)
+- [5. FAQ](#5-faq)
+- [引用](#引用)
+
+<a name="1"></a>
+## 1. 算法简介
+
+论文信息：
+> [TableMaster: PINGAN-VCGROUP’S SOLUTION FOR ICDAR 2021 COMPETITION ON SCIENTIFIC LITERATURE PARSING TASK B: TABLE RECOGNITION TO HTML](https://arxiv.org/pdf/2105.01848.pdf)
+> Ye, Jiaquan and Qi, Xianbiao and He, Yelin and Chen, Yihao and Gu, Dengyi and Gao, Peng and Xiao, Rong
+> 2021
+
+在PubTabNet表格识别公开数据集上，算法复现效果如下：
+
+|模型|骨干网络|配置文件|acc|下载链接|
+| --- | --- | --- | --- | --- |
+|TableMaster|TableResNetExtra|[configs/table/table_master.yml](../../configs/table/table_master.yml)|77.47%|[训练模型](https://paddleocr.bj.bcebos.com/ppstructure/models/tablemaster/table_structure_tablemaster_train.tar)/[推理模型](https://paddleocr.bj.bcebos.com/ppstructure/models/tablemaster/table_structure_tablemaster_infer.tar)|
+
+
+<a name="2"></a>
+## 2. 环境配置
+请先参考[《运行环境准备》](./environment.md)配置PaddleOCR运行环境，参考[《项目克隆》](./clone.md)克隆项目代码。
+
+
+<a name="3"></a>
+## 3. 模型训练、评估、预测
+
+上述TableMaster模型使用PubTabNet表格识别公开数据集训练得到，数据集下载可参考 [table_datasets](./dataset/table_datasets.md)。
+
+数据下载完成后，请参考[文本识别教程](./recognition.md)进行训练。PaddleOCR对代码进行了模块化，训练不同的模型只需要**更换配置文件**即可。
+
+<a name="4"></a>
+## 4. 推理部署
+
+<a name="4-1"></a>
+### 4.1 Python推理
+首先将训练得到best模型，转换成inference model。以基于TableResNetExtra骨干网络，在PubTabNet数据集训练的模型为例([模型下载地址](https://paddleocr.bj.bcebos.com/contribution/table_master.tar))，可以使用如下命令进行转换：
+
+```shell
+# 注意将pretrained_model的路径设置为本地路径。
+python3 tools/export_model.py -c configs/table/table_master.yml -o Global.pretrained_model=output/table_master/best_accuracy Global.save_inference_dir=./inference/table_master
+```
+
+**注意：**
+- 如果您是在自己的数据集上训练的模型，并且调整了字典文件，请注意修改配置文件中的`character_dict_path`是否为所正确的字典文件。
+
+转换成功后，在目录下有三个文件：
+```
+./inference/table_master/
+    ├── inference.pdiparams         # 识别inference模型的参数文件
+    ├── inference.pdiparams.info    # 识别inference模型的参数信息，可忽略
+    └── inference.pdmodel           # 识别inference模型的program文件
+```
+
+
+执行如下命令进行模型推理：
+
+```shell
+cd ppstructure/
+python3.7 table/predict_structure.py --table_model_dir=../output/table_master/table_structure_tablemaster_infer/ --table_algorithm=TableMaster --table_char_dict_path=../ppocr/utils/dict/table_master_structure_dict.txt --table_max_len=480 --image_dir=docs/table/table.jpg
+# 预测文件夹下所有图像时，可修改image_dir为文件夹，如 --image_dir='docs/table'。
+```
+
+执行命令后，上面图像的预测结果（结构信息和表格中每个单元格的坐标）会打印到屏幕上，同时会保存单元格坐标的可视化结果。示例如下：
+结果如下：
+```shell
+[2022/06/16 13:06:54] ppocr INFO: result: ['<html>', '<body>', '<table>', '<thead>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '</thead>', '<tbody>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '</tbody>', '</table>', '</body>', '</html>'], [[72.17591094970703, 10.759100914001465, 60.29658508300781, 16.6805362701416], [161.85562133789062, 10.884308815002441, 14.9495210647583, 16.727018356323242], [277.79876708984375, 29.54340362548828, 31.490320205688477, 18.143272399902344],
+...
+[336.11724853515625, 280.3601989746094, 39.456939697265625, 18.121286392211914]]
+[2022/06/16 13:06:54] ppocr INFO: save vis result to ./output/table.jpg
+[2022/06/16 13:06:54] ppocr INFO: Predict time of docs/table/table.jpg: 17.36806297302246
+```
+
+**注意**：
+
+- TableMaster在推理时比较慢，建议使用GPU进行使用。
+
+<a name="4-2"></a>
+### 4.2 C++推理部署
+
+由于C++预处理后处理还未支持TableMaster，所以暂未支持
+
+<a name="4-3"></a>
+### 4.3 Serving服务化部署
+
+暂不支持
+
+<a name="4-4"></a>
+### 4.4 更多推理部署
+
+暂不支持
+
+<a name="5"></a>
+## 5. FAQ
+
+## 引用
+
+```bibtex
+@article{ye2021pingan,
+  title={PingAn-VCGroup's Solution for ICDAR 2021 Competition on Scientific Literature Parsing Task B: Table Recognition to HTML},
+  author={Ye, Jiaquan and Qi, Xianbiao and He, Yelin and Chen, Yihao and Gu, Dengyi and Gao, Peng and Xiao, Rong},
+  journal={arXiv preprint arXiv:2105.01848},
+  year={2021}
+}
+```

doc/doc_ch/dataset/ocr_datasets.md

@@ -34,6 +34,7 @@ json.dumps编码前的图像标注信息是包含多个字典的list，字典中
 | ICDAR 2015 |https://rrc.cvc.uab.es/?ch=4&com=downloads| [train](https://paddleocr.bj.bcebos.com/dataset/train_icdar2015_label.txt) / [test](https://paddleocr.bj.bcebos.com/dataset/test_icdar2015_label.txt) |
 | ctw1500 |https://paddleocr.bj.bcebos.com/dataset/ctw1500.zip| 图片下载地址中已包含 |
 | total text |https://paddleocr.bj.bcebos.com/dataset/total_text.tar| 图片下载地址中已包含 |
+| td tr |https://paddleocr.bj.bcebos.com/dataset/TD_TR.tar| 图片下载地址中已包含 |
 
 #### 1.2.1 ICDAR 2015
 ICDAR 2015 数据集包含1000张训练图像和500张测试图像。ICDAR 2015 数据集可以从上表中链接下载，首次下载需注册。

doc/doc_ch/inference_ppocr.md

@@ -7,7 +7,8 @@
   - [1. 文本检测模型推理](#1-文本检测模型推理)
   - [2. 文本识别模型推理](#2-文本识别模型推理)
     - [2.1 超轻量中文识别模型推理](#21-超轻量中文识别模型推理)
-    - [2.2 多语言模型的推理](#22-多语言模型的推理)
+    - [2.2 英文识别模型推理](#22-英文识别模型推理)
+    - [2.3 多语言模型的推理](#23-多语言模型的推理)
   - [3. 方向分类模型推理](#3-方向分类模型推理)
   - [4. 文本检测、方向分类和文字识别串联推理](#4-文本检测方向分类和文字识别串联推理)
 
@@ -78,9 +79,29 @@ python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words/ch/word_4.jpg"
 Predicts of ./doc/imgs_words/ch/word_4.jpg:('实力活力', 0.9956803321838379)
 ```
 
+<a name="英文识别模型推理"></a>
+
+### 2.2 英文识别模型推理
+
+英文识别模型推理，可以执行如下命令， 注意修改字典路径：
+
+```
+# 下载英文数字识别模型：
+wget https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_det_infer.tar
+tar xf en_PP-OCRv3_det_infer.tar
+python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words/en/word_1.png" --rec_model_dir="./en_PP-OCRv3_det_infer/" --rec_char_dict_path="ppocr/utils/en_dict.txt"
+```
+
+![](../imgs_words/en/word_1.png)
+
+执行命令后，上图的预测结果为：
+
+```
+Predicts of ./doc/imgs_words/en/word_1.png: ('JOINT', 0.998160719871521)
+```
 <a name="多语言模型的推理"></a>
 
-### 2.2 多语言模型的推理
+### 2.3 多语言模型的推理
 
 如果您需要预测的是其他语言模型，可以在[此链接](./models_list.md#%E5%A4%9A%E8%AF%AD%E8%A8%80%E8%AF%86%E5%88%AB%E6%A8%A1%E5%9E%8B)中找到对应语言的inference模型，在使用inference模型预测时，需要通过`--rec_char_dict_path`指定使用的字典路径, 同时为了得到正确的可视化结果，需要通过 `--vis_font_path` 指定可视化的字体路径，`doc/fonts/` 路径下有默认提供的小语种字体，例如韩文识别：
 ```

doc/doc_en/algorithm_en.md

@@ -6,5 +6,6 @@ PaddleOCR will add cutting-edge OCR algorithms and models continuously. Check ou
 - [text detection algorithms](./algorithm_overview_en.md#11)
 - [text recognition algorithms](./algorithm_overview_en.md#12)
 - [end-to-end algorithms](./algorithm_overview_en.md#2)
+- [table recognition algorithms](./algorithm_overview_en.md#3)
 
 Developers are welcome to contribute more algorithms! Please refer to [add new algorithm](./add_new_algorithm_en.md) guideline.

doc/doc_en/algorithm_overview_en.md

 # OCR Algorithms
 
 - [1. Two-stage Algorithms](#1)
-  * [1.1 Text Detection Algorithms](#11)
-  * [1.2 Text Recognition Algorithms](#12)
+  - [1.1 Text Detection Algorithms](#11)
+  - [1.2 Text Recognition Algorithms](#12)
 - [2. End-to-end Algorithms](#2)
+- [3. Table Recognition Algorithms](#3)
 
 
 This tutorial lists the OCR algorithms supported by PaddleOCR, as well as the models and metrics of each algorithm on **English public datasets**. It is mainly used for algorithm introduction and algorithm performance comparison. For more models on other datasets including Chinese, please refer to [PP-OCR v2.0 models list](./models_list_en.md).
@@ -85,8 +86,9 @@ Refer to [DTRB](https://arxiv.org/abs/1904.01906), the training and evaluation r
 |SAR|Resnet31| 87.20% | rec_r31_sar | [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.1/rec/rec_r31_sar_train.tar) |
 |SEED|Aster_Resnet| 85.35% | rec_resnet_stn_bilstm_att | [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.1/rec/rec_resnet_stn_bilstm_att.tar) |
 |SVTR|SVTR-Tiny| 89.25% | rec_svtr_tiny_none_ctc_en | [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/rec_svtr_tiny_none_ctc_en_train.tar) |
-|ViTSTR|ViTSTR| 79.82% | rec_vitstr_none_ce_en | [trained model](https://paddleocr.bj.bcebos.com/rec_vitstr_none_none_train.tar) |
-|ABINet|Resnet45| 90.75% | rec_r45_abinet_en | [trained model](https://paddleocr.bj.bcebos.com/rec_r45_abinet_train.tar) |
+|ViTSTR|ViTSTR| 79.82% | rec_vitstr_none_ce | [trained model](https://paddleocr.bj.bcebos.com/rec_vitstr_none_none_train.tar) |
+|ABINet|Resnet45| 90.75% | rec_r45_abinet | [trained model](https://paddleocr.bj.bcebos.com/rec_r45_abinet_train.tar) |
+
 
 <a name="2"></a>
 
@@ -94,3 +96,15 @@ Refer to [DTRB](https://arxiv.org/abs/1904.01906), the training and evaluation r
 
 Supported end-to-end algorithms (Click the link to get the tutorial):
 - [x]  [PGNet](./algorithm_e2e_pgnet_en.md)
+
+<a name="3"></a>
+## 3. Table Recognition Algorithms
+
+Supported table recognition algorithms (Click the link to get the tutorial):
+- [x]  [TableMaster](./algorithm_table_master_en.md)
+
+On the PubTabNet dataset, the algorithm result is as follows:
+
+|Model|Backbone|Config|Acc|Download link|
+|---|---|---|---|---|
+|TableMaster|TableResNetExtra|[configs/table/table_master.yml](../../configs/table/table_master.yml)|77.47%|[trained](https://paddleocr.bj.bcebos.com/ppstructure/models/tablemaster/table_structure_tablemaster_train.tar) / [inference model](https://paddleocr.bj.bcebos.com/ppstructure/models/tablemaster/table_structure_tablemaster_infer.tar)|

doc/doc_en/algorithm_table_master_en.md

+# Table Recognition Algorithm-TableMASTER
+
+- [1. Introduction](#1-introduction)
+- [2. Environment](#2-environment)
+- [3. Model Training / Evaluation / Prediction](#3-model-training--evaluation--prediction)
+- [4. Inference and Deployment](#4-inference-and-deployment)
+  - [4.1 Python Inference](#41-python-inference)
+  - [4.2 C++ Inference](#42-c-inference)
+  - [4.3 Serving](#43-serving)
+  - [4.4 More](#44-more)
+- [5. FAQ](#5-faq)
+- [Citation](#citation)
+
+<a name="1"></a>
+## 1. Introduction
+
+Paper:
+> [TableMaster: PINGAN-VCGROUP’S SOLUTION FOR ICDAR 2021 COMPETITION ON SCIENTIFIC LITERATURE PARSING TASK B: TABLE RECOGNITION TO HTML](https://arxiv.org/pdf/2105.01848.pdf)
+> Ye, Jiaquan and Qi, Xianbiao and He, Yelin and Chen, Yihao and Gu, Dengyi and Gao, Peng and Xiao, Rong
+> 2021
+
+
+On the PubTabNet table recognition public data set, the algorithm reproduction acc is as follows:
+
+|Model|Backbone|Cnnfig|Acc|Download link|
+| --- | --- | --- | --- | --- |
+|TableMaster|TableResNetExtra|[configs/table/table_master.yml](../../configs/table/table_master.yml)|77.47%|[trained model](https://paddleocr.bj.bcebos.com/ppstructure/models/tablemaster/table_structure_tablemaster_train.tar)/[inference model](https://paddleocr.bj.bcebos.com/ppstructure/models/tablemaster/table_structure_tablemaster_infer.tar)|
+
+
+<a name="2"></a>
+## 2. Environment
+Please refer to ["Environment Preparation"](./environment_en.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone_en.md) to clone the project code.
+
+
+<a name="3"></a>
+## 3. Model Training / Evaluation / Prediction
+
+The above TableMaster model is trained using the PubTabNet table recognition public dataset. For the download of the dataset, please refer to [table_datasets](./dataset/table_datasets_en.md).
+
+After the data download is complete, please refer to [Text Recognition Training Tutorial](./recognition_en.md) for training. PaddleOCR has modularized the code structure, so that you only need to **replace the configuration file** to train different models.
+
+
+<a name="4"></a>
+## 4. Inference and Deployment
+
+<a name="4-1"></a>
+### 4.1 Python Inference
+
+First, convert the model saved in the TableMaster table recognition training process into an inference model. Taking the model based on the TableResNetExtra backbone network and trained on the PubTabNet dataset as example ([model download link](https://paddleocr.bj.bcebos.com/contribution/table_master.tar)), you can use the following command to convert:
+
+
+```shell
+python3 tools/export_model.py -c configs/table/table_master.yml -o Global.pretrained_model=output/table_master/best_accuracy Global.save_inference_dir=./inference/table_master
+```
+
+**Note: **
+- If you trained the model on your own dataset and adjusted the dictionary file, please pay attention to whether the `character_dict_path` in the modified configuration file is the correct dictionary file
+
+
+Execute the following command for model inference:
+
+```shell
+cd ppstructure/
+# When predicting all images in a folder, you can modify image_dir to a folder, such as --image_dir='docs/table'.
+python3.7 table/predict_structure.py --table_model_dir=../output/table_master/table_structure_tablemaster_infer/ --table_algorithm=TableMaster --table_char_dict_path=../ppocr/utils/dict/table_master_structure_dict.txt --table_max_len=480 --image_dir=docs/table/table.jpg
+
+```
+
+After executing the command, the prediction results of the above image (structural information and the coordinates of each cell in the table) are printed to the screen, and the visualization of the cell coordinates is also saved. An example is as follows:
+
+result：
+```shell
+[2022/06/16 13:06:54] ppocr INFO: result: ['<html>', '<body>', '<table>', '<thead>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '</thead>', '<tbody>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '</tbody>', '</table>', '</body>', '</html>'], [[72.17591094970703, 10.759100914001465, 60.29658508300781, 16.6805362701416], [161.85562133789062, 10.884308815002441, 14.9495210647583, 16.727018356323242], [277.79876708984375, 29.54340362548828, 31.490320205688477, 18.143272399902344],
+...
+[336.11724853515625, 280.3601989746094, 39.456939697265625, 18.121286392211914]]
+[2022/06/16 13:06:54] ppocr INFO: save vis result to ./output/table.jpg
+[2022/06/16 13:06:54] ppocr INFO: Predict time of docs/table/table.jpg: 17.36806297302246
+```
+
+**Note**:
+
+- TableMaster is relatively slow during inference, and it is recommended to use GPU for use.
+
+<a name="4-2"></a>
+### 4.2 C++ Inference
+
+Since the post-processing is not written in CPP, the TableMaster does not support CPP inference.
+
+
+<a name="4-3"></a>
+### 4.3 Serving
+
+Not supported
+
+<a name="4-4"></a>
+### 4.4 More
+
+Not supported
+
+<a name="5"></a>
+## 5. FAQ
+
+## Citation
+
+```bibtex
+@article{ye2021pingan,
+  title={PingAn-VCGroup's Solution for ICDAR 2021 Competition on Scientific Literature Parsing Task B: Table Recognition to HTML},
+  author={Ye, Jiaquan and Qi, Xianbiao and He, Yelin and Chen, Yihao and Gu, Dengyi and Gao, Peng and Xiao, Rong},
+  journal={arXiv preprint arXiv:2105.01848},
+  year={2021}
+}
+```

doc/doc_en/inference_ppocr_en.md

@@ -8,7 +8,8 @@ This article introduces the use of the Python inference engine for the PP-OCR mo
   - [Text Detection Model Inference](#text-detection-model-inference)
   - [Text Recognition Model Inference](#text-recognition-model-inference)
     - [1. Lightweight Chinese Recognition Model Inference](#1-lightweight-chinese-recognition-model-inference)
-    - [2. Multilingual Model Inference](#2-multilingual-model-inference)
+    - [2. English Recognition Model Inference](#2-english-recognition-model-inference)
+    - [3. Multilingual Model Inference](#3-multilingual-model-inference)
   - [Angle Classification Model Inference](#angle-classification-model-inference)
   - [Text Detection Angle Classification and Recognition Inference Concatenation](#text-detection-angle-classification-and-recognition-inference-concatenation)
 
@@ -76,10 +77,31 @@ After executing the command, the prediction results (recognized text and score)
 ```bash
 Predicts of ./doc/imgs_words_en/word_10.png:('PAIN', 0.988671)
 ```
+<a name="2-english-recognition-model-inference"></a>
+### 2. English Recognition Model Inference
 
-<a name="MULTILINGUAL_MODEL_INFERENCE"></a>
+For English recognition model inference, you can execute the following commands,you need to specify the dictionary path used by `--rec_char_dict_path`:
 
-### 2. Multilingual Model Inference
+```
+# download en model：
+wget https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_det_infer.tar
+tar xf en_PP-OCRv3_det_infer.tar
+python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words/en/word_1.png" --rec_model_dir="./en_PP-OCRv3_det_infer/" --rec_char_dict_path="ppocr/utils/en_dict.txt"
+```
+
+![](../imgs_words/en/word_1.png)
+
+
+After executing the command, the prediction result of the above figure is:
+
+```
+Predicts of ./doc/imgs_words/en/word_1.png: ('JOINT', 0.998160719871521)
+```
+
+
+<a name="3-multilingual-model-inference"></a>
+
+### 3. Multilingual Model Inference
 If you need to predict [other language models](./models_list_en.md#Multilingual), when using inference model prediction, you need to specify the dictionary path used by `--rec_char_dict_path`. At the same time, in order to get the correct visualization results,
 You need to specify the visual font path through `--vis_font_path`. There are small language fonts provided by default under the `doc/fonts` path, such as Korean recognition:
 

ppocr/data/imaug/__init__.py

@@ -23,7 +23,8 @@ from .random_crop_data import EastRandomCropData, RandomCropImgMask
 from .make_pse_gt import MakePseGt
 
 
-from .rec_img_aug import RecAug, RecConAug, RecResizeImg, ClsResizeImg, \
+
+from .rec_img_aug import BaseDataAugmentation, RecAug, RecConAug, RecResizeImg, ClsResizeImg, \
                          SRNRecResizeImg, GrayRecResizeImg, SARRecResizeImg, PRENResizeImg, \
                          ABINetRecResizeImg, SVTRRecResizeImg, ABINetRecAug
 from .ssl_img_aug import SSLRotateResize
@@ -36,7 +37,7 @@ from .label_ops import *
 from .east_process import *
 from .sast_process import *
 from .pg_process import *
-from .gen_table_mask import *
+from .table_ops import *
 
 from .vqa import *
 

ppocr/data/imaug/operators.py

@@ -205,9 +205,12 @@ class DetResizeForTest(object):
     def __init__(self, **kwargs):
         super(DetResizeForTest, self).__init__()
         self.resize_type = 0
+        self.keep_ratio = False
         if 'image_shape' in kwargs:
             self.image_shape = kwargs['image_shape']
             self.resize_type = 1
+            if 'keep_ratio' in kwargs:
+                self.keep_ratio = kwargs['keep_ratio']
         elif 'limit_side_len' in kwargs:
             self.limit_side_len = kwargs['limit_side_len']
             self.limit_type = kwargs.get('limit_type', 'min')
@@ -237,6 +240,10 @@ class DetResizeForTest(object):
     def resize_image_type1(self, img):
         resize_h, resize_w = self.image_shape
         ori_h, ori_w = img.shape[:2]  # (h, w, c)
+        if self.keep_ratio is True:
+            resize_w = ori_w * resize_h / ori_h
+            N = math.ceil(resize_w / 32)
+            resize_w = N * 32
         ratio_h = float(resize_h) / ori_h
         ratio_w = float(resize_w) / ori_w
         img = cv2.resize(img, (int(resize_w), int(resize_h)))

ppocr/data/imaug/gen_table_mask.py → ppocr/data/imaug/table_ops.py

@@ -48,10 +48,12 @@ class GenTableMask(object):
         in_text = False  # 是否遍历到了字符区内
         box_list = []
         for i in range(len(project_val_array)):
-            if in_text == False and project_val_array[i] > spilt_threshold:  # 进入字符区了
+            if in_text == False and project_val_array[
+                    i] > spilt_threshold:  # 进入字符区了
                 in_text = True
                 start_idx = i
-            elif project_val_array[i] <= spilt_threshold and in_text == True:  # 进入空白区了
+            elif project_val_array[
+                    i] <= spilt_threshold and in_text == True:  # 进入空白区了
                 end_idx = i
                 in_text = False
                 if end_idx - start_idx <= 2:
@@ -70,7 +72,8 @@ class GenTableMask(object):
         box_gray_img = cv2.cvtColor(box_img, cv2.COLOR_BGR2GRAY)
         h, w = box_gray_img.shape
         # 灰度图片进行二值化处理
-        ret, thresh1 = cv2.threshold(box_gray_img, 200, 255, cv2.THRESH_BINARY_INV)
+        ret, thresh1 = cv2.threshold(box_gray_img, 200, 255,
+                                     cv2.THRESH_BINARY_INV)
         # 纵向腐蚀
         if h < w:
             kernel = np.ones((2, 1), np.uint8)
@@ -95,10 +98,12 @@ class GenTableMask(object):
         box_list = []
         spilt_threshold = 0
         for i in range(len(project_val_array)):
-            if in_text == False and project_val_array[i] > spilt_threshold:  # 进入字符区了
+            if in_text == False and project_val_array[
+                    i] > spilt_threshold:  # 进入字符区了
                 in_text = True
                 start_idx = i
-            elif project_val_array[i] <= spilt_threshold and in_text == True:  # 进入空白区了
+            elif project_val_array[
+                    i] <= spilt_threshold and in_text == True:  # 进入空白区了
                 end_idx = i
                 in_text = False
                 if end_idx - start_idx <= 2:
@@ -120,7 +125,8 @@ class GenTableMask(object):
                     h_end = h
                 word_img = erosion[h_start:h_end + 1, :]
                 word_h, word_w = word_img.shape
-                w_split_list, w_projection_map = self.projection(word_img.T, word_w, word_h)
+                w_split_list, w_projection_map = self.projection(word_img.T,
+                                                                 word_w, word_h)
                 w_start, w_end = w_split_list[0][0], w_split_list[-1][1]
                 if h_start > 0:
                     h_start -= 1
@@ -170,7 +176,8 @@ class GenTableMask(object):
 
                 for sno in range(len(split_bbox_list)):
                     left, top, right, bottom = split_bbox_list[sno]
-                    left, top, right, bottom = self.shrink_bbox([left, top, right, bottom])
+                    left, top, right, bottom = self.shrink_bbox(
+                        [left, top, right, bottom])
                     if self.mask_type == 1:
                         mask_img[top:bottom, left:right] = 1.0
                         data['mask_img'] = mask_img
@@ -179,66 +186,44 @@ class GenTableMask(object):
                         data['image'] = mask_img
         return data
 
+
 class ResizeTableImage(object):
-    def __init__(self, max_len, **kwargs):
+    def __init__(self, max_len, resize_bboxes=False, infer_mode=False,
+                 **kwargs):
         super(ResizeTableImage, self).__init__()
         self.max_len = max_len
+        self.resize_bboxes = resize_bboxes
+        self.infer_mode = infer_mode
 
-    def get_img_bbox(self, cells):
-        bbox_list = []
-        if len(cells) == 0:
-            return bbox_list
-        cell_num = len(cells)
-        for cno in range(cell_num):
-            if "bbox" in cells[cno]:
-                bbox = cells[cno]['bbox']
-                bbox_list.append(bbox)
-        return bbox_list
-
-    def resize_img_table(self, img, bbox_list, max_len):
+    def __call__(self, data):
+        img = data['image']
         height, width = img.shape[0:2]
-        ratio = max_len / (max(height, width) * 1.0)
+        ratio = self.max_len / (max(height, width) * 1.0)
         resize_h = int(height * ratio)
         resize_w = int(width * ratio)
-        img_new = cv2.resize(img, (resize_w, resize_h))
-        bbox_list_new = []
-        for bno in range(len(bbox_list)):
-            left, top, right, bottom = bbox_list[bno].copy()
-            left = int(left * ratio)
-            top = int(top * ratio)
-            right = int(right * ratio)
-            bottom = int(bottom * ratio)
-            bbox_list_new.append([left, top, right, bottom])
-        return img_new, bbox_list_new
-    
-    def __call__(self, data):
-        img = data['image']
-        if 'cells' not in data:
-            cells = []
-        else:
-            cells = data['cells']
-        bbox_list = self.get_img_bbox(cells)
-        img_new, bbox_list_new = self.resize_img_table(img, bbox_list, self.max_len)
-        data['image'] = img_new
-        cell_num = len(cells)
-        bno = 0
-        for cno in range(cell_num):
-            if "bbox" in data['cells'][cno]:
-                data['cells'][cno]['bbox'] = bbox_list_new[bno]
-                bno += 1
+        resize_img = cv2.resize(img, (resize_w, resize_h))
+        if self.resize_bboxes and not self.infer_mode:
+            data['bboxes'] = data['bboxes'] * ratio
+        data['image'] = resize_img
+        data['src_img'] = img
+        data['shape'] = np.array([resize_h, resize_w, ratio, ratio])
         data['max_len'] = self.max_len
         return data
 
+
 class PaddingTableImage(object):
-    def __init__(self, **kwargs):
+    def __init__(self, size, **kwargs):
         super(PaddingTableImage, self).__init__()
+        self.size = size
 
     def __call__(self, data):
         img = data['image']
-        max_len = data['max_len']
-        padding_img = np.zeros((max_len, max_len, 3), dtype=np.float32)
+        pad_h, pad_w = self.size
+        padding_img = np.zeros((pad_h, pad_w, 3), dtype=np.float32)
         height, width = img.shape[0:2]
         padding_img[0:height, 0:width, :] = img.copy()
         data['image'] = padding_img
+        shape = data['shape'].tolist()
+        shape.extend([pad_h, pad_w])
+        data['shape'] = np.array(shape)
         return data
-            
\ No newline at end of file

ppocr/data/imaug/vqa/__init__.py

@@ -13,7 +13,12 @@
 # limitations under the License.
 
 from .token import VQATokenPad, VQASerTokenChunk, VQAReTokenChunk, VQAReTokenRelation
+from .augment import DistortBBox
 
 __all__ = [
-    'VQATokenPad', 'VQASerTokenChunk', 'VQAReTokenChunk', 'VQAReTokenRelation'
+    'VQATokenPad',
+    'VQASerTokenChunk',
+    'VQAReTokenChunk',
+    'VQAReTokenRelation',
+    'DistortBBox',
 ]

ppocr/data/imaug/vqa/augment.py

+# copyright (c) 2022 PaddlePaddle Authors. All Rights Reserve.
+#
+# 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 os
+import sys
+import numpy as np
+import random
+
+
+class DistortBBox:
+    def __init__(self, prob=0.5, max_scale=1, **kwargs):
+        """Random distort bbox
+        """
+        self.prob = prob
+        self.max_scale = max_scale
+
+    def __call__(self, data):
+        if random.random() > self.prob:
+            return data
+        bbox = np.array(data['bbox'])
+        rnd_scale = (np.random.rand(*bbox.shape) - 0.5) * 2 * self.max_scale
+        bbox = np.round(bbox + rnd_scale).astype(bbox.dtype)
+        data['bbox'] = np.clip(data['bbox'], 0, 1000)
+        data['bbox'] = bbox.tolist()
+        sys.stdout.flush()
+        return data