[Feature] Support lightweight model for Japanese recognition (#548)

README.md

@@ -28,8 +28,9 @@
 ## 🥳 What's New
 
 - Aug. 2024:
-  - ✨✨✨ Support [Segment-Anything-2](https://docs.ultralytics.com/tasks/pose/) model! (Recommended)
   - 🤗 Pre-release the latest version [2.4.0](https://github.com/CVHub520/X-AnyLabeling/releases/tag/v2.4.0) 🤗
+  - ✨✨✨ Support [Segment-Anything-2](https://github.com/facebookresearch/segment-anything-2) model! (Recommended)
+  - 👏👏👏 Support lightweight model for Japanese recognition.
 - Jul. 2024:
   - Add PPOCR-Recognition and KIE import/export functionality for training PP-OCR task.
   - Add ODVG import/export functionality for training grounding task.
@@ -56,8 +57,8 @@
   - Support for quick tag correction is available; please refer to this [document](./docs/en/user_guide.md) for guidance.
   - Release version [2.3.1](https://github.com/CVHub520/X-AnyLabeling/releases/tag/v2.3.1).
 - Jan. 2024:
-  - 👏👏👏 Combining CLIP and SAM models for enhanced semantic and spatial understanding. An example can be found [here](./anylabeling/configs/auto_labeling/edge_sam_with_chinese_clip.yaml).
-  - 🔥🔥🔥 Adding support for the [Depth Anything](https://github.com/LiheYoung/Depth-Anything.git) model in the depth estimation task.
+  - Combining CLIP and SAM models for enhanced semantic and spatial understanding. An example can be found [here](./anylabeling/configs/auto_labeling/edge_sam_with_chinese_clip.yaml).
+  - Add support for the [Depth Anything](https://github.com/LiheYoung/Depth-Anything.git) model in the depth estimation task.
   - Release version [2.3.0](https://github.com/CVHub520/X-AnyLabeling/releases/tag/v2.3.0).
   - Support [YOLOv8-OBB](https://github.com/ultralytics/ultralytics) model.
   - Support [RTMDet](https://github.com/open-mmlab/mmyolo/tree/main/configs/rtmdet) and [RTMO](https://github.com/open-mmlab/mmpose/tree/main/projects/rtmpose) model.

README_zh-CN.md

@@ -46,7 +46,8 @@
 
 - 2024年8月:
   - 🤗 预发布[X-AnyLabeling v2.4.0](https://github.com/CVHub520/X-AnyLabeling/releases/tag/v2.4.0)最新版本 🤗
-  - ✨✨✨ 支持[Segment-Anything-2](https://docs.ultralytics.com/tasks/pose/)模型。
+  - ✨✨✨ 支持[Segment-Anything-2](https://github.com/facebookresearch/segment-anything-2)模型。
+  - 👏👏👏 支持[日文字符识别](./anylabeling/configs/auto_labeling/japan_ppocr.yaml)模型。
 - 2024年7月:
   - 新增 PPOCR 识别和关键信息提取标签导入/导出功能。
   - 新增 ODVG 标签导入/导出功能，以支持 Grounding 模型训练。

anylabeling/configs/auto_labeling/ch_ppocr_v4.yaml

@@ -5,4 +5,5 @@ det_model_path: https://github.com/CVHub520/X-AnyLabeling/releases/download/v0.2
 rec_model_path: https://github.com/CVHub520/X-AnyLabeling/releases/download/v0.2.2/ch_PP-OCRv4_rec_infer.onnx
 cls_model_path: https://github.com/CVHub520/X-AnyLabeling/releases/download/v0.2.2/ch_ppocr_mobile_v2.0_cls_infer.onnx
 drop_score: 0.5
+lang: ch
 use_angle_cls: True

anylabeling/configs/auto_labeling/japan_ppocr.yaml

@@ -0,0 +1,9 @@
+type: ppocr_v4
+name: japan_ppocr-r20240803
+display_name: japan_PP-OCRv3 (PaddleOCR)
+det_model_path: https://github.com/CVHub520/X-AnyLabeling/releases/download/v0.2.2/ch_PP-OCRv4_det_infer.onnx
+rec_model_path: https://github.com/CVHub520/X-AnyLabeling/releases/download/v2.4.0/japan_PP-OCRv3_rec_infer.onnx
+cls_model_path: https://github.com/CVHub520/X-AnyLabeling/releases/download/v0.2.2/ch_ppocr_mobile_v2.0_cls_infer.onnx
+drop_score: 0.5
+lang: japan
+use_angle_cls: True

anylabeling/configs/auto_labeling/models.yaml

@@ -1,5 +1,7 @@
 - model_name: "ch_ppocr_v4-r20230915"
   config_file: ":/ch_ppocr_v4.yaml"
+- model_name: "japan_ppocr-r20240803"
+  config_file: ":/japan_ppocr.yaml"
 - model_name: "clrnet_tusimple_r18-r20230901"
   config_file: ":/clrnet_tusimple_r18.yaml"
 - model_name: "damo_yolo_l-r20231001"