use rapidocr

machine-learning/immich_ml/config.py

@@ -41,6 +41,7 @@ class PreloadModelData(BaseModel):
 
 class MaxBatchSize(BaseModel):
     facial_recognition: int | None = None
+    text_recognition: int | None = None
 
 
 class Settings(BaseSettings):

machine-learning/immich_ml/models/__init__.py

@@ -3,12 +3,14 @@ from typing import Any
 from immich_ml.models.base import InferenceModel
 from immich_ml.models.clip.textual import MClipTextualEncoder, OpenClipTextualEncoder
 from immich_ml.models.clip.visual import OpenClipVisualEncoder
+from immich_ml.models.ocr.detection import TextDetector
+from immich_ml.models.ocr.recognition import TextRecognizer
 from immich_ml.schemas import ModelSource, ModelTask, ModelType
 
 from .constants import get_model_source
 from .facial_recognition.detection import FaceDetector
 from .facial_recognition.recognition import FaceRecognizer
-from .ocr.paddle import PaddleOCRecognizer
+
 
 def get_model_class(model_name: str, model_type: ModelType, model_task: ModelTask) -> type[InferenceModel]:
     source = get_model_source(model_name)
@@ -28,8 +30,11 @@ def get_model_class(model_name: str, model_type: ModelType, model_task: ModelTas
         case ModelSource.INSIGHTFACE, ModelType.RECOGNITION, ModelTask.FACIAL_RECOGNITION:
             return FaceRecognizer
 
-        case ModelSource.PADDLE, ModelType.OCR, ModelTask.OCR:
+        case ModelSource.PADDLE, ModelType.DETECTION, ModelTask.OCR:
-            return PaddleOCRecognizer
+            return TextDetector
+
+        case ModelSource.PADDLE, ModelType.RECOGNITION, ModelTask.OCR:
+            return TextRecognizer
 
         case _:
             raise ValueError(f"Unknown model combination: {source}, {model_type}, {model_task}")

machine-learning/immich_ml/models/base.py

@@ -38,9 +38,8 @@ class InferenceModel(ABC):
 
     def download(self) -> None:
         if not self.cached:
-            log.info(
+            model_type = self.model_type.replace("-", " ")
-                f"Downloading {self.model_type.replace('-', ' ')} model '{self.model_name}'. This may take a while."
+            log.info(f"Downloading {model_type} model '{self.model_name}' to {self.model_path}. This may take a while.")
-            )
             self._download()
 
     def load(self) -> None:

machine-learning/immich_ml/models/ocr/detection.py

@@ -0,0 +1,79 @@
+from typing import Any
+
+import numpy as np
+from PIL import Image
+from rapidocr.ch_ppocr_det import TextDetector as RapidTextDetector
+from rapidocr.inference_engine.base import FileInfo, InferSession
+from rapidocr.utils import DownloadFile, DownloadFileInput
+from rapidocr.utils.typings import EngineType, LangDet, OCRVersion, TaskType
+from rapidocr.utils.typings import ModelType as RapidModelType
+
+from immich_ml.config import log
+from immich_ml.models.base import InferenceModel
+from immich_ml.models.transforms import decode_cv2
+from immich_ml.schemas import ModelSession, ModelTask, ModelType
+
+from .schemas import OcrOptions, TextDetectionOutput
+
+
+class TextDetector(InferenceModel):
+    depends = []
+    identity = (ModelType.DETECTION, ModelTask.OCR)
+
+    def __init__(self, model_name: str, **model_kwargs: Any) -> None:
+        super().__init__(model_name, **model_kwargs)
+        self.max_resolution = 1440
+        self.min_score = 0.5
+        self.score_mode = "fast"
+        self._empty: TextDetectionOutput = {
+            "resized": np.empty(0, dtype=np.float32),
+            "boxes": np.empty(0, dtype=np.float32),
+            "scores": (),
+        }
+
+    def _download(self) -> None:
+        model_info = InferSession.get_model_url(
+            FileInfo(
+                engine_type=EngineType.ONNXRUNTIME,
+                ocr_version=OCRVersion.PPOCRV5,
+                task_type=TaskType.DET,
+                lang_type=LangDet.CH,
+                model_type=RapidModelType.MOBILE if "mobile" in self.model_name else RapidModelType.SERVER,
+            )
+        )
+        download_params = DownloadFileInput(
+            file_url=model_info["model_dir"],
+            sha256=model_info["SHA256"],
+            save_path=self.model_path,
+            logger=log,
+        )
+        DownloadFile.run(download_params)
+
+    def _load(self) -> ModelSession:
+        session = self._make_session(self.model_path)
+        self.model = RapidTextDetector(
+            OcrOptions(
+                session=session.session,
+                limit_side_len=self.max_resolution,
+                limit_type="min",
+                box_thresh=self.min_score,
+                score_mode=self.score_mode,
+            )
+        )
+        return session
+
+    def configure(self, **kwargs: Any) -> None:
+        self.max_resolution = kwargs.get("maxResolution", self.max_resolution)
+        self.min_score = kwargs.get("minScore", self.min_score)
+        self.score_mode = kwargs.get("scoreMode", self.score_mode)
+
+    def _predict(self, inputs: bytes | Image.Image, **kwargs: Any) -> TextDetectionOutput:
+        results = self.model(decode_cv2(inputs))
+        if results.boxes is None or results.scores is None or results.img is None:
+            return self._empty
+        log.info(f"{results.boxes=}, {results.scores=}")
+        return {
+            "resized": results.img,
+            "boxes": np.array(results.boxes, dtype=np.float32),
+            "scores": np.array(results.scores, dtype=np.float32),
+        }

machine-learning/immich_ml/models/ocr/paddle.py

@@ -1,51 +0,0 @@
-from typing import Any, List
-
-import numpy as np
-from numpy.typing import NDArray
-from paddleocr import PaddleOCR
-from PIL import Image
-from immich_ml.models.base import InferenceModel
-from immich_ml.models.transforms import decode_cv2
-from immich_ml.schemas import OCROutput, ModelTask, ModelType
-
-class PaddleOCRecognizer(InferenceModel):
-    depends = []
-    identity = (ModelType.OCR, ModelTask.OCR)
-
-    def __init__(self, model_name: str, **model_kwargs: Any) -> None:
-        self.orientation_classify_enabled = model_kwargs.get("orientationClassifyEnabled", False)
-        self.unwarping_enabled = model_kwargs.get("unwarpingEnabled", False)
-        super().__init__(model_name, **model_kwargs)
-        self._load()
-        self.loaded = True
-
-    def _load(self) -> PaddleOCR:
-        self.model = PaddleOCR(
-            text_detection_model_name=f"{self.model_name}_det",
-            text_recognition_model_name=f"{self.model_name}_rec",
-            use_doc_orientation_classify=self.orientation_classify_enabled,
-            use_doc_unwarping=self.unwarping_enabled,
-        )
-
-    def configure(self, **kwargs: Any) -> None:
-        self.min_detection_score = kwargs.get("minDetectionScore", 0.3)
-        self.min_detection_box_score = kwargs.get("minDetectionBoxScore", 0.6)
-        self.min_recognition_score = kwargs.get("minRecognitionScore", 0.0)
-
-    def _predict(self, inputs: NDArray[np.uint8] | bytes | Image.Image, **kwargs: Any) -> List[OCROutput]:
-        inputs = decode_cv2(inputs)
-        results = self.model.predict(
-            inputs,
-            text_det_thresh=self.min_detection_score,
-            text_det_box_thresh=self.min_detection_box_score,
-            text_rec_score_thresh=self.min_recognition_score
-        )
-        return [
-            OCROutput(
-                text=text, confidence=score,
-                x1=box[0][0], y1=box[0][1], x2=box[1][0], y2=box[1][1],
-                x3=box[2][0], y3=box[2][1], x4=box[3][0], y4=box[3][1]
-            )
-            for result in results
-            for text, score, box in zip(result['rec_texts'], result['rec_scores'], result['rec_polys'])
-        ]

machine-learning/immich_ml/models/ocr/recognition.py

@@ -0,0 +1,115 @@
+from typing import Any
+
+import cv2
+import numpy as np
+from PIL.Image import Image
+from rapidocr.ch_ppocr_rec import TextRecInput
+from rapidocr.ch_ppocr_rec import TextRecognizer as RapidTextRecognizer
+from rapidocr.inference_engine.base import FileInfo, InferSession
+from rapidocr.utils import DownloadFile, DownloadFileInput
+from rapidocr.utils.typings import EngineType, LangDet, OCRVersion, TaskType
+from rapidocr.utils.typings import ModelType as RapidModelType
+
+from immich_ml.config import log, settings
+from immich_ml.models.base import InferenceModel
+from immich_ml.schemas import ModelSession, ModelTask, ModelType
+
+from .schemas import OcrOptions, TextDetectionOutput, TextRecognitionOutput
+
+
+class TextRecognizer(InferenceModel):
+    depends = [(ModelType.DETECTION, ModelTask.OCR)]
+    identity = (ModelType.RECOGNITION, ModelTask.OCR)
+
+    def __init__(self, model_name: str, **model_kwargs: Any) -> None:
+        self.min_score = model_kwargs.get("minScore", 0.5)
+        self._empty: TextRecognitionOutput = {
+            "box": np.empty(0, dtype=np.float32),
+            "boxScore": [],
+            "text": [],
+            "textScore": [],
+        }
+        super().__init__(model_name, **model_kwargs)
+
+    def _download(self) -> None:
+        model_info = InferSession.get_model_url(
+            FileInfo(
+                engine_type=EngineType.ONNXRUNTIME,
+                ocr_version=OCRVersion.PPOCRV5,
+                task_type=TaskType.REC,
+                lang_type=LangDet.CH,
+                model_type=RapidModelType.MOBILE if "mobile" in self.model_name else RapidModelType.SERVER,
+            )
+        )
+        download_params = DownloadFileInput(
+            file_url=model_info["model_dir"],
+            sha256=model_info["SHA256"],
+            save_path=self.model_path,
+            logger=log,
+        )
+        DownloadFile.run(download_params)
+
+    def _load(self) -> ModelSession:
+        session = self._make_session(self.model_path)
+        self.model = RapidTextRecognizer(
+            OcrOptions(
+                session=session.session,
+                rec_batch_num=settings.max_batch_size.text_recognition if settings.max_batch_size is not None else 6,
+                rec_img_shape=(3, 48, 320),
+            )
+        )
+        return session
+
+    def configure(self, **kwargs: Any) -> None:
+        self.min_score = kwargs.get("minScore", self.min_score)
+
+    def _predict(self, _: Image, texts: TextDetectionOutput, **kwargs: Any) -> TextRecognitionOutput:
+        boxes, resized_img, box_scores = texts["boxes"], texts["resized"], texts["scores"]
+        if boxes.shape[0] == 0:
+            return self._empty
+        rec = self.model(TextRecInput(img=self.get_crop_img_list(resized_img, boxes)))
+        if rec.txts is None:
+            return self._empty
+
+        height, width = resized_img.shape[0:2]
+        log.info(f"Image shape: width={width}, height={height}")
+        boxes[:, :, 0] /= width
+        boxes[:, :, 1] /= height
+
+        text_scores = np.array(rec.scores)
+        valid_text_score_idx = text_scores > 0.5
+        valid_score_idx_list = valid_text_score_idx.tolist()
+        return {
+            "box": boxes.reshape(-1, 8)[valid_text_score_idx],
+            "text": [rec.txts[i] for i in range(len(rec.txts)) if valid_score_idx_list[i]],
+            "boxScore": box_scores[valid_text_score_idx],
+            "textScore": text_scores[valid_text_score_idx],
+        }
+
+    def get_crop_img_list(self, img: np.ndarray, boxes: np.ndarray) -> list[np.ndarray]:
+        img_crop_width = np.maximum(
+            np.linalg.norm(boxes[:, 1] - boxes[:, 0], axis=1), np.linalg.norm(boxes[:, 2] - boxes[:, 3], axis=1)
+        ).astype(np.int32)
+        img_crop_height = np.maximum(
+            np.linalg.norm(boxes[:, 0] - boxes[:, 3], axis=1), np.linalg.norm(boxes[:, 1] - boxes[:, 2], axis=1)
+        ).astype(np.int32)
+        pts_std = np.zeros((img_crop_width.shape[0], 4, 2), dtype=np.float32)
+        pts_std[:, 1:3, 0] = img_crop_width[:, None]
+        pts_std[:, 2:4, 1] = img_crop_height[:, None]
+
+        img_crop_sizes = np.stack([img_crop_width, img_crop_height], axis=1).tolist()
+        imgs = []
+        for box, pts_std, dst_size in zip(list(boxes), list(pts_std), img_crop_sizes):
+            M = cv2.getPerspectiveTransform(box, pts_std)
+            dst_img = cv2.warpPerspective(
+                img,
+                M,
+                dst_size,
+                borderMode=cv2.BORDER_REPLICATE,
+                flags=cv2.INTER_CUBIC,
+            )
+            dst_height, dst_width = dst_img.shape[0:2]
+            if dst_height * 1.0 / dst_width >= 1.5:
+                dst_img = np.rot90(dst_img)
+            imgs.append(dst_img)
+        return imgs

machine-learning/immich_ml/models/ocr/schemas.py

@@ -0,0 +1,26 @@
+from typing import Iterable
+
+import numpy as np
+import numpy.typing as npt
+from rapidocr.utils.typings import EngineType
+from typing_extensions import TypedDict
+
+
+class TextDetectionOutput(TypedDict):
+    resized: npt.NDArray[np.float32]
+    boxes: npt.NDArray[np.float32]
+    scores: Iterable[float]
+
+
+class TextRecognitionOutput(TypedDict):
+    box: npt.NDArray[np.float32]
+    boxScore: Iterable[float]
+    text: Iterable[str]
+    textScore: Iterable[float]
+
+
+# RapidOCR expects engine_type to be an attribute
+class OcrOptions(dict):
+    def __init__(self, **options):
+        super().__init__(**options)
+        self.engine_type = EngineType.ONNXRUNTIME

machine-learning/immich_ml/schemas.py

@@ -25,6 +25,7 @@ class ModelTask(StrEnum):
     SEARCH = "clip"
     OCR = "ocr"
 
+
 class ModelType(StrEnum):
     DETECTION = "detection"
     RECOGNITION = "recognition"
@@ -32,6 +33,7 @@ class ModelType(StrEnum):
     VISUAL = "visual"
     OCR = "ocr"
 
+
 class ModelFormat(StrEnum):
     ARMNN = "armnn"
     ONNX = "onnx"
@@ -44,6 +46,7 @@ class ModelSource(StrEnum):
     OPENCLIP = "openclip"
     PADDLE = "paddle"
 
+
 ModelIdentity = tuple[ModelType, ModelTask]
 
 
@@ -87,19 +90,6 @@ class DetectedFace(TypedDict):
 FacialRecognitionOutput = list[DetectedFace]
 
 
-class OCROutput(TypedDict):
-    text: str
-    confidence: float
-    x1: int
-    y1: int
-    x2: int
-    y2: int
-    x3: int
-    y3: int
-    x4: int
-    y4: int
-
-
 class PipelineEntry(TypedDict):
     modelName: str
     options: dict[str, Any]

machine-learning/pyproject.toml

@@ -24,6 +24,7 @@ dependencies = [
     "uvicorn[standard]>=0.22.0,<1.0",
     "paddleocr>=3.0.0",
     "setuptools>=78.1.0",
+    "rapidocr>=3.1.0",
 ]
 
 [dependency-groups]
@@ -50,15 +51,11 @@ lint = [
 dev = ["locust>=2.15.1", { include-group = "test" }, { include-group = "lint" }]
 
 [project.optional-dependencies]
-cpu = ["onnxruntime>=1.15.0,<2", "paddlepaddle==3.0.0rc1"]
+cpu = ["onnxruntime>=1.15.0,<2"]
-cuda = ["onnxruntime-gpu>=1.17.0,<2", "paddlepaddle-gpu==3.0.0rc1"]
+cuda = ["onnxruntime-gpu>=1.17.0,<2"]
-openvino = ["onnxruntime-openvino>=1.17.1,<1.19.0", "paddlepaddle==3.0.0rc1"]
+openvino = ["onnxruntime-openvino>=1.17.1,<1.19.0"]
-armnn = ["onnxruntime>=1.15.0,<2", "paddlepaddle==3.0.0rc1"]
+armnn = ["onnxruntime>=1.15.0,<2"]
-rknn = [
+rknn = ["onnxruntime>=1.15.0,<2", "rknn-toolkit-lite2>=2.3.0,<3"]
-    "onnxruntime>=1.15.0,<2",
-    "rknn-toolkit-lite2>=2.3.0,<3",
-    "paddlepaddle==3.0.0rc1",
-]
 rocm = []
 
 [tool.uv]

machine-learning/uv.lock

@@ -71,6 +71,12 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/78/b6/6307fbef88d9b5ee7421e68d78a9f162e0da4900bc5f5793f6d3d0e34fb8/annotated_types-0.7.0-py3-none-any.whl", hash = "sha256:1f02e8b43a8fbbc3f3e0d4f0f4bfc8131bcb4eebe8849b8e5c773f3a1c582a53", size = 13643 },
 ]
 
+[[package]]
+name = "antlr4-python3-runtime"
+version = "4.9.3"
+source = { registry = "https://pypi.org/simple" }
+sdist = { url = "https://files.pythonhosted.org/packages/3e/38/7859ff46355f76f8d19459005ca000b6e7012f2f1ca597746cbcd1fbfe5e/antlr4-python3-runtime-4.9.3.tar.gz", hash = "sha256:f224469b4168294902bb1efa80a8bf7855f24c99aef99cbefc1bcd3cce77881b", size = 117034 }
+
 [[package]]
 name = "anyio"
 version = "4.2.0"
@@ -86,15 +92,6 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/bf/cd/d6d9bb1dadf73e7af02d18225cbd2c93f8552e13130484f1c8dcfece292b/anyio-4.2.0-py3-none-any.whl", hash = "sha256:745843b39e829e108e518c489b31dc757de7d2131d53fac32bd8df268227bfee", size = 85481 },
 ]
 
-[[package]]
-name = "astor"
-version = "0.8.1"
-source = { registry = "https://pypi.org/simple" }
-sdist = { url = "https://files.pythonhosted.org/packages/5a/21/75b771132fee241dfe601d39ade629548a9626d1d39f333fde31bc46febe/astor-0.8.1.tar.gz", hash = "sha256:6a6effda93f4e1ce9f618779b2dd1d9d84f1e32812c23a29b3fff6fd7f63fa5e", size = 35090 }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/c3/88/97eef84f48fa04fbd6750e62dcceafba6c63c81b7ac1420856c8dcc0a3f9/astor-0.8.1-py2.py3-none-any.whl", hash = "sha256:070a54e890cefb5b3739d19f30f5a5ec840ffc9c50ffa7d23cc9fc1a38ebbfc5", size = 27488 },
-]
-
 [[package]]
 name = "backports-asyncio-runner"
 version = "1.2.0"
@@ -671,15 +668,6 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/e3/7f/f584f5d15323feb897d42ef0e9d910649e2150d7a30cf7e7a8cc1d236e6f/Cython-3.0.8-py2.py3-none-any.whl", hash = "sha256:171b27051253d3f9108e9759e504ba59ff06e7f7ba944457f94deaf9c21bf0b6", size = 1168213 },
 ]
 
-[[package]]
-name = "decorator"
-version = "5.2.1"
-source = { registry = "https://pypi.org/simple" }
-sdist = { url = "https://files.pythonhosted.org/packages/43/fa/6d96a0978d19e17b68d634497769987b16c8f4cd0a7a05048bec693caa6b/decorator-5.2.1.tar.gz", hash = "sha256:65f266143752f734b0a7cc83c46f4618af75b8c5911b00ccb61d0ac9b6da0360", size = 56711 }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/4e/8c/f3147f5c4b73e7550fe5f9352eaa956ae838d5c51eb58e7a25b9f3e2643b/decorator-5.2.1-py3-none-any.whl", hash = "sha256:d316bb415a2d9e2d2b3abcc4084c6502fc09240e292cd76a76afc106a1c8e04a", size = 9190 },
-]
-
 [[package]]
 name = "easydict"
 version = "1.11"
@@ -1168,6 +1156,7 @@ dependencies = [
     { name = "pydantic" },
     { name = "pydantic-settings" },
     { name = "python-multipart" },
+    { name = "rapidocr" },
     { name = "rich" },
     { name = "setuptools" },
     { name = "tokenizers" },
@@ -1177,23 +1166,18 @@ dependencies = [
 [package.optional-dependencies]
 armnn = [
     { name = "onnxruntime" },
-    { name = "paddlepaddle" },
 ]
 cpu = [
     { name = "onnxruntime" },
-    { name = "paddlepaddle" },
 ]
 cuda = [
     { name = "onnxruntime-gpu" },
-    { name = "paddlepaddle-gpu" },
 ]
 openvino = [
     { name = "onnxruntime-openvino" },
-    { name = "paddlepaddle" },
 ]
 rknn = [
     { name = "onnxruntime" },
-    { name = "paddlepaddle" },
     { name = "rknn-toolkit-lite2" },
 ]
 
@@ -1256,15 +1240,11 @@ requires-dist = [
     { name = "opencv-python-headless", specifier = ">=4.7.0.72,<5.0" },
     { name = "orjson", specifier = ">=3.9.5" },
     { name = "paddleocr", specifier = ">=3.0.0" },
-    { name = "paddlepaddle", marker = "extra == 'armnn'", specifier = "==3.0.0rc1", index = "https://www.paddlepaddle.org.cn/packages/stable/cpu/" },
-    { name = "paddlepaddle", marker = "extra == 'cpu'", specifier = "==3.0.0rc1", index = "https://www.paddlepaddle.org.cn/packages/stable/cpu/" },
-    { name = "paddlepaddle", marker = "extra == 'openvino'", specifier = "==3.0.0rc1", index = "https://www.paddlepaddle.org.cn/packages/stable/cpu/" },
-    { name = "paddlepaddle", marker = "extra == 'rknn'", specifier = "==3.0.0rc1", index = "https://www.paddlepaddle.org.cn/packages/stable/cpu/" },
-    { name = "paddlepaddle-gpu", marker = "extra == 'cuda'", specifier = "==3.0.0rc1", index = "https://www.paddlepaddle.org.cn/packages/stable/cu118/" },
     { name = "pillow", specifier = ">=9.5.0,<11.0" },
     { name = "pydantic", specifier = ">=2.0.0,<3" },
     { name = "pydantic-settings", specifier = ">=2.5.2,<3" },
     { name = "python-multipart", specifier = ">=0.0.6,<1.0" },
+    { name = "rapidocr", specifier = ">=3.1.0" },
     { name = "rich", specifier = ">=13.4.2" },
     { name = "rknn-toolkit-lite2", marker = "extra == 'rknn'", specifier = ">=2.3.0,<3" },
     { name = "setuptools", specifier = ">=78.1.0" },
@@ -1836,107 +1816,16 @@ wheels = [
 ]
 
 [[package]]
-name = "nvidia-cublas-cu11"
+name = "omegaconf"
-version = "11.11.3.6"
+version = "2.3.0"
-source = { registry = "https://pypi.org/simple" }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/46/be/c222e33e60d28ecd496a46fc4d78ccae0ee28e1fd7dc705b6288b4cad27e/nvidia_cublas_cu11-11.11.3.6-py3-none-manylinux1_x86_64.whl", hash = "sha256:39fb40e8f486dd8a2ddb8fdeefe1d5b28f5b99df01c87ab3676f057a74a5a6f3", size = 417870452 },
-    { url = "https://files.pythonhosted.org/packages/ea/2e/9d99c60771d275ecf6c914a612e9a577f740a615bc826bec132368e1d3ae/nvidia_cublas_cu11-11.11.3.6-py3-none-manylinux2014_x86_64.whl", hash = "sha256:60252822adea5d0b10cd990a7dc7bedf7435f30ae40083c7a624a85a43225abc", size = 417870460 },
-]
-
-[[package]]
-name = "nvidia-cuda-cupti-cu11"
-version = "11.8.87"
-source = { registry = "https://pypi.org/simple" }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/27/c9/b4b15f709a694ea9f84871c6c4fbeeb54bab225962d852665a2c6f77f90d/nvidia_cuda_cupti_cu11-11.8.87-py3-none-manylinux1_x86_64.whl", hash = "sha256:0e50c707df56c75a2c0703dc6b886f3c97a22f37d6f63839f75b7418ba672a8d", size = 13093657 },
-    { url = "https://files.pythonhosted.org/packages/74/42/9f5c5cc084ce6f3073048c4f6806f45ba4c8c73f227c9587215d9c372e05/nvidia_cuda_cupti_cu11-11.8.87-py3-none-manylinux2014_x86_64.whl", hash = "sha256:4191a17913a706b5098681280cd089cd7d8d3df209a6f5cb79384974a96d24f2", size = 13093662 },
-]
-
-[[package]]
-name = "nvidia-cuda-nvrtc-cu11"
-version = "11.8.89"
-source = { registry = "https://pypi.org/simple" }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/83/08/a9833e4e9f9165bedb7f36033b47aa399b053b9cb2eaf7b84d1e28705cf7/nvidia_cuda_nvrtc_cu11-11.8.89-py3-none-manylinux1_x86_64.whl", hash = "sha256:1f27d67b0f72902e9065ae568b4f6268dfe49ba3ed269c9a3da99bb86d1d2008", size = 23173264 },
-    { url = "https://files.pythonhosted.org/packages/60/44/202e027c224c26e15a53f01c5c7604c7f6b4fd368882d3164ea08fead207/nvidia_cuda_nvrtc_cu11-11.8.89-py3-none-manylinux2014_x86_64.whl", hash = "sha256:a8d02f3cba345be56b1ffc3e74d8f61f02bb758dd31b0f20e12277a5a244f756", size = 23173745 },
-]
-
-[[package]]
-name = "nvidia-cuda-runtime-cu11"
-version = "11.8.89"
-source = { registry = "https://pypi.org/simple" }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/45/3e/84db02be49fe6d6df6e42f69fd64501c22d0f9ada9c9877f885612085d20/nvidia_cuda_runtime_cu11-11.8.89-py3-none-manylinux1_x86_64.whl", hash = "sha256:f587bd726eb2f7612cf77ce38a2c1e65cf23251ff49437f6161ce0d647f64f7c", size = 875585 },
-    { url = "https://files.pythonhosted.org/packages/a6/ec/a540f28b31de7bc1ed49eecc72035d4cb77db88ead1d42f7bfa5ae407ac6/nvidia_cuda_runtime_cu11-11.8.89-py3-none-manylinux2014_x86_64.whl", hash = "sha256:92d04069a987e1fbc9213f8376d265df0f7bb42617d44f5eda1f496acea7f2d1", size = 875592 },
-]
-
-[[package]]
-name = "nvidia-cudnn-cu11"
-version = "8.9.6.50"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
-    { name = "nvidia-cublas-cu11", marker = "(platform_machine != 'aarch64' and sys_platform == 'linux') or (sys_platform != 'darwin' and sys_platform != 'linux')" },
+    { name = "antlr4-python3-runtime" },
-    { name = "nvidia-cuda-nvrtc-cu11", marker = "(platform_machine != 'aarch64' and sys_platform == 'linux') or (sys_platform != 'darwin' and sys_platform != 'linux')" },
+    { name = "pyyaml" },
 ]
+sdist = { url = "https://files.pythonhosted.org/packages/09/48/6388f1bb9da707110532cb70ec4d2822858ddfb44f1cdf1233c20a80ea4b/omegaconf-2.3.0.tar.gz", hash = "sha256:d5d4b6d29955cc50ad50c46dc269bcd92c6e00f5f90d23ab5fee7bfca4ba4cc7", size = 3298120 }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/85/2d/3f083fcff1c302119f48e7b30a5f7b23db793f262f900943a9eb456b9e4d/nvidia_cudnn_cu11-8.9.6.50-py3-none-manylinux1_x86_64.whl", hash = "sha256:319a8f7ca3d65139f1b69998595c7076ae0e4271a325e5dfde50a3ca31f55584", size = 699874407 },
+    { url = "https://files.pythonhosted.org/packages/e3/94/1843518e420fa3ed6919835845df698c7e27e183cb997394e4a670973a65/omegaconf-2.3.0-py3-none-any.whl", hash = "sha256:7b4df175cdb08ba400f45cae3bdcae7ba8365db4d165fc65fd04b050ab63b46b", size = 79500 },
-]
-
-[[package]]
-name = "nvidia-cufft-cu11"
-version = "10.9.0.58"
-source = { registry = "https://pypi.org/simple" }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/74/79/b912a77e38e41f15a0581a59f5c3548d1ddfdda3225936fb67c342719e7a/nvidia_cufft_cu11-10.9.0.58-py3-none-manylinux1_x86_64.whl", hash = "sha256:222f9da70c80384632fd6035e4c3f16762d64ea7a843829cb278f98b3cb7dd81", size = 168405414 },
-    { url = "https://files.pythonhosted.org/packages/64/c8/133717b43182ba063803e983e7680a94826a9f4ff5734af0ca315803f1b3/nvidia_cufft_cu11-10.9.0.58-py3-none-manylinux2014_x86_64.whl", hash = "sha256:e21037259995243cc370dd63c430d77ae9280bedb68d5b5a18226bfc92e5d748", size = 168405419 },
-]
-
-[[package]]
-name = "nvidia-curand-cu11"
-version = "10.3.0.86"
-source = { registry = "https://pypi.org/simple" }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/49/28/c47f8e2439ddbcbeae3cf74d43ed572b651d630ea72863d5357f3759eb66/nvidia_curand_cu11-10.3.0.86-py3-none-manylinux1_x86_64.whl", hash = "sha256:ac439548c88580269a1eb6aeb602a5aed32f0dbb20809a31d9ed7d01d77f6bf5", size = 58124493 },
-    { url = "https://files.pythonhosted.org/packages/58/e5/ce5806afc48a6e4e0dddd25316ac60b6fa94fd1791bdbf4ca17bf52696ea/nvidia_curand_cu11-10.3.0.86-py3-none-manylinux2014_x86_64.whl", hash = "sha256:cd4cffbf78bb06580206b4814d5dc696d1161c902aae37b2bba00056832379e6", size = 58124497 },
-]
-
-[[package]]
-name = "nvidia-cusolver-cu11"
-version = "11.4.1.48"
-source = { registry = "https://pypi.org/simple" }
-dependencies = [
-    { name = "nvidia-cublas-cu11", marker = "(platform_machine != 'aarch64' and sys_platform == 'linux') or (sys_platform != 'darwin' and sys_platform != 'linux')" },
-]
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/55/ee/939ff0104991dd7bdabb4c9767994c612ba0e1c9a55672a1ddd42f5e5b16/nvidia_cusolver_cu11-11.4.1.48-py3-none-manylinux1_x86_64.whl", hash = "sha256:ca538f545645b7e6629140786d3127fe067b3d5a085bd794cde5bfe877c8926f", size = 128240842 },
-    { url = "https://files.pythonhosted.org/packages/52/fe/866e87e6e6a1b0a5fcf8524a058042656702f2057e22bfdb8899a7c38e10/nvidia_cusolver_cu11-11.4.1.48-py3-none-manylinux2014_x86_64.whl", hash = "sha256:ea9fb1ad8c644ca9ed55af13cc39af3b7ba4c3eb5aef18471fe1fe77d94383cb", size = 128246438 },
-]
-
-[[package]]
-name = "nvidia-cusparse-cu11"
-version = "11.7.5.86"
-source = { registry = "https://pypi.org/simple" }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/c1/e0/21b829c535d569831835a4ca5d049a19ba00d3e91f3e12ab4ad27bd7385f/nvidia_cusparse_cu11-11.7.5.86-py3-none-manylinux1_x86_64.whl", hash = "sha256:4ae709fe78d3f23f60acaba8c54b8ad556cf16ca486e0cc1aa92dca7555d2d2b", size = 204126221 },
-    { url = "https://files.pythonhosted.org/packages/ed/5c/b0333b07c51ced77397c2fb0d9826072cea0da9d421aa7e792aa0f8ecc72/nvidia_cusparse_cu11-11.7.5.86-py3-none-manylinux2014_x86_64.whl", hash = "sha256:8d7cf1628fd8d462b5d2ba6678fae34733a48ecb80495b9c68672ec6a6dde5ef", size = 204126227 },
-]
-
-[[package]]
-name = "nvidia-nccl-cu11"
-version = "2.19.3"
-source = { registry = "https://pypi.org/simple" }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/0e/7d/cc3dbf36c5af39b042d508b7a441ada1fce69bd18c800e5c25dc4e9f8933/nvidia_nccl_cu11-2.19.3-py3-none-manylinux1_x86_64.whl", hash = "sha256:7c58afbeddf7f7c6b7dd7d84a7f4e85462610ee0c656287388b96d89dcf046d5", size = 135288005 },
-]
-
-[[package]]
-name = "nvidia-nvtx-cu11"
-version = "11.8.86"
-source = { registry = "https://pypi.org/simple" }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/d5/a2/23214c23118784dc2189ac2d2e48190df3e4206e2f73eb17d47140797a2b/nvidia_nvtx_cu11-11.8.86-py3-none-manylinux1_x86_64.whl", hash = "sha256:890656d8bd9b4e280231c832e1f0d03459200ba4824ddda3dcb59b1e1989b9f5", size = 99125 },
-    { url = "https://files.pythonhosted.org/packages/b5/ad/973a187b137a3d45dc3faac421ef1275fb41fc169fd3889e2d5ceb0daa54/nvidia_nvtx_cu11-11.8.86-py3-none-manylinux2014_x86_64.whl", hash = "sha256:979f5b2aef5da164c5c53c64c85c3dfa61b8b4704f4f963bb568bf98fa8472e8", size = 99130 },
 ]
 
 [[package]]
@@ -2059,6 +1948,23 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/a7/9e/7110d2c5d543ab03b9581dbb1f8e2429863e44e0c9b4960b766f230c1279/opencv_contrib_python-4.10.0.84-cp37-abi3-win_amd64.whl", hash = "sha256:47ec3160dae75f70e099b286d1a2e086d20dac8b06e759f60eaf867e6bdecba7", size = 45541421 },
 ]
 
+[[package]]
+name = "opencv-python"
+version = "4.11.0.86"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "numpy" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/17/06/68c27a523103dad5837dc5b87e71285280c4f098c60e4fe8a8db6486ab09/opencv-python-4.11.0.86.tar.gz", hash = "sha256:03d60ccae62304860d232272e4a4fda93c39d595780cb40b161b310244b736a4", size = 95171956 }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/05/4d/53b30a2a3ac1f75f65a59eb29cf2ee7207ce64867db47036ad61743d5a23/opencv_python-4.11.0.86-cp37-abi3-macosx_13_0_arm64.whl", hash = "sha256:432f67c223f1dc2824f5e73cdfcd9db0efc8710647d4e813012195dc9122a52a", size = 37326322 },
+    { url = "https://files.pythonhosted.org/packages/3b/84/0a67490741867eacdfa37bc18df96e08a9d579583b419010d7f3da8ff503/opencv_python-4.11.0.86-cp37-abi3-macosx_13_0_x86_64.whl", hash = "sha256:9d05ef13d23fe97f575153558653e2d6e87103995d54e6a35db3f282fe1f9c66", size = 56723197 },
+    { url = "https://files.pythonhosted.org/packages/f3/bd/29c126788da65c1fb2b5fb621b7fed0ed5f9122aa22a0868c5e2c15c6d23/opencv_python-4.11.0.86-cp37-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:1b92ae2c8852208817e6776ba1ea0d6b1e0a1b5431e971a2a0ddd2a8cc398202", size = 42230439 },
+    { url = "https://files.pythonhosted.org/packages/2c/8b/90eb44a40476fa0e71e05a0283947cfd74a5d36121a11d926ad6f3193cc4/opencv_python-4.11.0.86-cp37-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:6b02611523803495003bd87362db3e1d2a0454a6a63025dc6658a9830570aa0d", size = 62986597 },
+    { url = "https://files.pythonhosted.org/packages/fb/d7/1d5941a9dde095468b288d989ff6539dd69cd429dbf1b9e839013d21b6f0/opencv_python-4.11.0.86-cp37-abi3-win32.whl", hash = "sha256:810549cb2a4aedaa84ad9a1c92fbfdfc14090e2749cedf2c1589ad8359aa169b", size = 29384337 },
+    { url = "https://files.pythonhosted.org/packages/a4/7d/f1c30a92854540bf789e9cd5dde7ef49bbe63f855b85a2e6b3db8135c591/opencv_python-4.11.0.86-cp37-abi3-win_amd64.whl", hash = "sha256:085ad9b77c18853ea66283e98affefe2de8cc4c1f43eda4c100cf9b2721142ec", size = 39488044 },
+]
+
 [[package]]
 name = "opencv-python-headless"
 version = "4.11.0.86"
@@ -2076,18 +1982,6 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/86/8a/69176a64335aed183529207ba8bc3d329c2999d852b4f3818027203f50e6/opencv_python_headless-4.11.0.86-cp37-abi3-win_amd64.whl", hash = "sha256:6c304df9caa7a6a5710b91709dd4786bf20a74d57672b3c31f7033cc638174ca", size = 39402386 },
 ]
 
-[[package]]
-name = "opt-einsum"
-version = "3.3.0"
-source = { registry = "https://pypi.org/simple" }
-dependencies = [
-    { name = "numpy" },
-]
-sdist = { url = "https://files.pythonhosted.org/packages/7d/bf/9257e53a0e7715bc1127e15063e831f076723c6cd60985333a1c18878fb8/opt_einsum-3.3.0.tar.gz", hash = "sha256:59f6475f77bbc37dcf7cd748519c0ec60722e91e63ca114e68821c0c54a46549", size = 73951 }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/bc/19/404708a7e54ad2798907210462fd950c3442ea51acc8790f3da48d2bee8b/opt_einsum-3.3.0-py3-none-any.whl", hash = "sha256:2455e59e3947d3c275477df7f5205b30635e266fe6dc300e3d9f9646bfcea147", size = 65486 },
-]
-
 [[package]]
 name = "orjson"
 version = "3.11.3"
@@ -2188,74 +2082,6 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/74/72/c8218cc7489762ab3d1fb25721f8c928f10085e38feb8d52fd6583bfc592/paddleocr-3.2.0-py3-none-any.whl", hash = "sha256:2b942295ad5963de8e01d68afb15a9507d713bc7299e2dfeb198d9c3ac5cf76f", size = 75976 },
 ]
 
-[[package]]
-name = "paddlepaddle"
-version = "3.0.0rc1"
-source = { registry = "https://www.paddlepaddle.org.cn/packages/stable/cpu/" }
-dependencies = [
-    { name = "astor" },
-    { name = "decorator" },
-    { name = "httpx" },
-    { name = "networkx" },
-    { name = "numpy" },
-    { name = "opt-einsum" },
-    { name = "pillow" },
-    { name = "protobuf" },
-    { name = "typing-extensions" },
-]
-wheels = [
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp310-cp310-linux_aarch64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp310-cp310-linux_x86_64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp310-cp310-macosx_10_9_x86_64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp310-cp310-macosx_11_0_arm64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp310-cp310-win_amd64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp311-cp311-linux_aarch64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp311-cp311-linux_x86_64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp311-cp311-macosx_10_9_x86_64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp311-cp311-macosx_11_0_arm64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp311-cp311-win_amd64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp312-cp312-linux_aarch64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp312-cp312-linux_x86_64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp312-cp312-macosx_10_9_x86_64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp312-cp312-macosx_11_0_arm64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cpu/paddlepaddle/paddlepaddle-3.0.0rc1-cp312-cp312-win_amd64.whl" },
-]
-
-[[package]]
-name = "paddlepaddle-gpu"
-version = "3.0.0rc1"
-source = { registry = "https://www.paddlepaddle.org.cn/packages/stable/cu118/" }
-dependencies = [
-    { name = "astor" },
-    { name = "decorator" },
-    { name = "httpx" },
-    { name = "networkx" },
-    { name = "numpy" },
-    { name = "nvidia-cublas-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "nvidia-cuda-cupti-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "nvidia-cuda-nvrtc-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "nvidia-cuda-runtime-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "nvidia-cudnn-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "nvidia-cufft-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "nvidia-curand-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "nvidia-cusolver-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "nvidia-cusparse-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "nvidia-nccl-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "nvidia-nvtx-cu11", marker = "platform_machine == 'x86_64' and sys_platform == 'linux'" },
-    { name = "opt-einsum" },
-    { name = "pillow" },
-    { name = "protobuf" },
-    { name = "typing-extensions" },
-]
-wheels = [
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cu118/paddlepaddle-gpu/paddlepaddle_gpu-3.0.0rc1-cp310-cp310-linux_x86_64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cu118/paddlepaddle-gpu/paddlepaddle_gpu-3.0.0rc1-cp310-cp310-win_amd64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cu118/paddlepaddle-gpu/paddlepaddle_gpu-3.0.0rc1-cp311-cp311-linux_x86_64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cu118/paddlepaddle-gpu/paddlepaddle_gpu-3.0.0rc1-cp311-cp311-win_amd64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cu118/paddlepaddle-gpu/paddlepaddle_gpu-3.0.0rc1-cp312-cp312-linux_x86_64.whl" },
-    { url = "https://paddle-whl.bj.bcebos.com/stable/cu118/paddlepaddle-gpu/paddlepaddle_gpu-3.0.0rc1-cp312-cp312-win_amd64.whl" },
-]
-
 [[package]]
 name = "paddlex"
 version = "3.2.1"
@@ -2967,6 +2793,27 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/f0/a1/a5f4bebaa31d109003909809d88aeb0d4b201463a9ea29308d9e4f9e7655/qudida-0.0.4-py3-none-any.whl", hash = "sha256:4519714c40cd0f2e6c51e1735edae8f8b19f4efe1f33be13e9d644ca5f736dd6", size = 3478 },
 ]
 
+[[package]]
+name = "rapidocr"
+version = "3.4.0"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "colorlog" },
+    { name = "numpy" },
+    { name = "omegaconf" },
+    { name = "opencv-python" },
+    { name = "pillow" },
+    { name = "pyclipper" },
+    { name = "pyyaml" },
+    { name = "requests" },
+    { name = "shapely" },
+    { name = "six" },
+    { name = "tqdm" },
+]
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/e1/e4/09ec2657421f1f23eec6b40ecbee77bbf3ff053c1483d8f2ed62d285bcf3/rapidocr-3.4.0-py3-none-any.whl", hash = "sha256:08d72f4c3a566bc76ac5c8d65d1e1c39550222b3b41b73aef976914ce80f48db", size = 15055924 },
+]
+
 [[package]]
 name = "requests"
 version = "2.32.3"