add ocr detV5 models (#3976)

paddlex/configs/modules/text_detection/PP-OCRv5_mobile_det.yaml

@@ -0,0 +1,40 @@
+Global:
+  model: PP-OCRv5_mobile_det
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  module: text_det
+  dataset_dir: "/paddle/dataset/paddlex/ocr_det/ocr_det_dataset_examples"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert:
+    enable: False
+    src_dataset_type: null
+  split:
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  epochs_iters: 100
+  batch_size: 4
+  learning_rate: 0.001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_mobile_det_pretrained.pdparams
+  resume_path: null
+  log_interval: 10
+  eval_interval: 1
+  save_interval: 1
+
+Evaluate:
+  weight_path: "output/best_accuracy/best_accuracy.pdparams"
+  log_interval: 1
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_mobile_det_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_accuracy/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_001.png"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/text_detection/PP-OCRv5_server_det.yaml

@@ -0,0 +1,40 @@
+Global:
+  model: PP-OCRv5_server_det
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  module: text_det
+  dataset_dir: "/paddle/dataset/paddlex/ocr_det/ocr_det_dataset_examples"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert:
+    enable: False
+    src_dataset_type: null
+  split:
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  epochs_iters: 100
+  batch_size: 4
+  learning_rate: 0.001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_server_det_pretrained.pdparams
+  resume_path: null
+  log_interval: 10
+  eval_interval: 1
+  save_interval: 1
+
+Evaluate:
+  weight_path: "output/best_accuracy/best_accuracy.pdparams"
+  log_interval: 1
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_server_det_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_accuracy/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_001.png"
+  kernel_option:
+    run_mode: paddle

paddlex/inference/utils/official_models.py

@@ -338,6 +338,8 @@ PP-LCNet_x1_0_vehicle_attribute_infer.tar",
     "PP-DocBee-2B": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/PP-DocBee-2B_infer.tar",
     "PP-DocBee-7B": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/PP-DocBee-7B_infer.tar",
     "PP-Chart2Table": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/PP-Chart2Table_infer.tar",
+    "PP-OCRv5_server_det": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/PP-OCRv5_server_det_infer.tar",
+    "PP-OCRv5_mobile_det": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/PP-OCRv5_mobile_det_infer.tar",
     "PP-OCRv5_server_rec": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/\
 PP-OCRv5_server_rec_infer.tar",
 }

paddlex/modules/text_detection/model_list.py

@@ -13,6 +13,8 @@
 # limitations under the License.
 
 MODELS = [
+    "PP-OCRv5_mobile_det",
+    "PP-OCRv5_server_det",
     "PP-OCRv4_mobile_det",
     "PP-OCRv4_server_det",
     "PP-OCRv4_mobile_seal_det",

paddlex/repo_apis/PaddleOCR_api/configs/PP-OCRv5_mobile_det.yaml

@@ -0,0 +1,171 @@
+Global:
+  debug: false
+  use_gpu: true
+  epoch_num: &epoch_num 100
+  log_smooth_window: 20
+  print_batch_step: 10
+  save_model_dir: output
+  save_epoch_step: 1
+  eval_batch_step:
+  - 0
+  - 100
+  cal_metric_during_train: false
+  checkpoints:
+  pretrained_model: https://paddleocr.bj.bcebos.com/pretrained/PPLCNetV3_x0_75_ocr_det.pdparams
+  save_inference_dir: null
+  use_visualdl: false
+  distributed: true
+  d2s_train_image_shape: [3, 640, 640]
+
+Architecture:
+  model_type: det
+  algorithm: DB
+  Transform: null
+  Backbone:
+    name: PPLCNetV3
+    scale: 0.75
+    det: True
+  Neck:
+    name: RSEFPN
+    out_channels: 96
+    shortcut: True
+  Head:
+    name: DBHead
+    k: 50
+    fix_nan: True
+
+Loss:
+  name: DBLoss
+  balance_loss: true
+  main_loss_type: DiceLoss
+  alpha: 5
+  beta: 10
+  ohem_ratio: 3
+
+Optimizer:
+  name: Adam
+  beta1: 0.9
+  beta2: 0.999
+  lr:
+    name: Cosine
+    learning_rate: 0.001 
+    warmup_epoch: 2
+  regularizer:
+    name: L2
+    factor: 1e-6
+
+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: TextDetDataset
+    data_dir: datasets/ICDAR2015
+    label_file_list:
+      - datasets/ICDAR2015/train.txt
+    transforms:
+    - DecodeImage:
+        img_mode: BGR
+        channel_first: false
+    - DetLabelEncode: null
+    - CopyPaste: 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: 50
+        keep_ratio: true
+    - MakeBorderMap:
+        shrink_ratio: 0.4
+        thresh_min: 0.3
+        thresh_max: 0.7
+        total_epoch: *epoch_num
+    - MakeShrinkMap:
+        shrink_ratio: 0.4
+        min_text_size: 8
+        total_epoch: *epoch_num
+    - NormalizeImage:
+        scale: 1./255.
+        mean:
+        - 0.485
+        - 0.456
+        - 0.406
+        std:
+        - 0.229
+        - 0.224
+        - 0.225
+        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: 8
+    num_workers: 8
+
+Eval:
+  dataset:
+    name: TextDetDataset
+    data_dir: datasets/ICDAR2015
+    label_file_list:
+      - datasets/ICDAR2015/val.txt
+    transforms:
+    - DecodeImage:
+        img_mode: BGR
+        channel_first: false
+    - DetLabelEncode: null
+    - DetResizeForTest:
+        resize_long: 960
+    - NormalizeImage:
+        scale: 1./255.
+        mean:
+        - 0.485
+        - 0.456
+        - 0.406
+        std:
+        - 0.229
+        - 0.224
+        - 0.225
+        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: 0
+profiler_options: null

paddlex/repo_apis/PaddleOCR_api/configs/PP-OCRv5_server_det.yaml

@@ -0,0 +1,171 @@
+Global:
+  debug: false
+  use_gpu: true
+  epoch_num: &epoch_num 100
+  log_smooth_window: 20
+  print_batch_step: 10
+  save_model_dir: output
+  save_epoch_step: 1
+  eval_batch_step:
+  - 0
+  - 100
+  cal_metric_during_train: false
+  checkpoints:
+  pretrained_model: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PPHGNetV2_B4_ocr_det.pdparams
+  save_inference_dir: null
+  use_visualdl: false
+  distributed: true
+  d2s_train_image_shape: [3, 640, 640]
+
+Architecture:
+  model_type: det
+  algorithm: DB
+  Transform: null
+  Backbone:
+    name: PPHGNetV2_B4
+    det: True
+  Neck:
+    name: LKPAN
+    out_channels: 256
+    intracl: true
+  Head:
+    name: PFHeadLocal
+    k: 50
+    mode: "large"
+    
+
+Loss:
+  name: DBLoss
+  balance_loss: true
+  main_loss_type: DiceLoss
+  alpha: 5
+  beta: 10
+  ohem_ratio: 3
+
+Optimizer:
+  name: Adam
+  beta1: 0.9
+  beta2: 0.999
+  lr:
+    name: Cosine
+    learning_rate: 0.001 
+    warmup_epoch: 2
+  regularizer:
+    name: L2
+    factor: 1e-6
+
+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: TextDetDataset
+    data_dir: datasets/ICDAR2015
+    label_file_list:
+      - datasets/ICDAR2015/train.txt
+    transforms:
+    - DecodeImage:
+        img_mode: BGR
+        channel_first: false
+    - DetLabelEncode: null
+    - CopyPaste: 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: 50
+        keep_ratio: true
+    - MakeBorderMap:
+        shrink_ratio: 0.4
+        thresh_min: 0.3
+        thresh_max: 0.7
+        total_epoch: *epoch_num
+    - MakeShrinkMap:
+        shrink_ratio: 0.4
+        min_text_size: 8
+        total_epoch: *epoch_num
+    - NormalizeImage:
+        scale: 1./255.
+        mean:
+        - 0.485
+        - 0.456
+        - 0.406
+        std:
+        - 0.229
+        - 0.224
+        - 0.225
+        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: 3
+
+Eval:
+  dataset:
+    name: TextDetDataset
+    data_dir: datasets/ICDAR2015
+    label_file_list:
+      - datasets/ICDAR2015/val.txt
+    transforms:
+    - DecodeImage:
+        img_mode: BGR
+        channel_first: false
+    - DetLabelEncode: null
+    - DetResizeForTest:
+        resize_long: 960
+    - NormalizeImage:
+        scale: 1./255.
+        mean:
+        - 0.485
+        - 0.456
+        - 0.406
+        std:
+        - 0.229
+        - 0.224
+        - 0.225
+        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: 0
+profiler_options: null

paddlex/repo_apis/PaddleOCR_api/text_det/register.py

@@ -87,3 +87,21 @@ register_model_info(
         "supported_apis": ["train", "evaluate", "predict", "export"],
     }
 )
+
+register_model_info(
+    {
+        "model_name": "PP-OCRv5_server_det",
+        "suite": "TextDet",
+        "config_path": osp.join(PDX_CONFIG_DIR, "PP-OCRv5_server_det.yaml"),
+        "supported_apis": ["train", "evaluate", "predict", "export"],
+    }
+)
+
+register_model_info(
+    {
+        "model_name": "PP-OCRv5_mobile_det",
+        "suite": "TextDet",
+        "config_path": osp.join(PDX_CONFIG_DIR, "PP-OCRv5_mobile_det.yaml"),
+        "supported_apis": ["train", "evaluate", "predict", "export"],
+    }
+)