deepdoctection 0.39.7__py3-none-any.whl → 0.41.0__py3-none-any.whl

deepdoctection/pipe/text.py

@@ -70,7 +70,6 @@ class TextExtractionService(PipelineComponent):
         text_extract_detector: Union[ObjectDetector, PdfMiner, TextRecognizer],
         extract_from_roi: Optional[Union[Sequence[TypeOrStr], TypeOrStr]] = None,
         run_time_ocr_language_selection: bool = False,
-        skip_if_text_extracted: bool = False,
     ):
         """
         :param text_extract_detector: ObjectDetector
@@ -79,8 +78,6 @@ class TextExtractionService(PipelineComponent):
                                                 multiple language selections. Also requires that a language detection
                                                 pipeline component ran before. It will select the expert language OCR
                                                 model based on the determined language.
-        :param skip_if_text_extracted: Set to `True` if text has already been extracted in a previous pipeline component
-                                       and should not be extracted again. Use-case: A PDF with some scanned images.
         """
 
         if extract_from_roi is None:
@@ -104,11 +101,6 @@ class TextExtractionService(PipelineComponent):
                 raise TypeError("Only TesseractOcrDetector supports multiple languages")
 
         self.run_time_ocr_language_selection = run_time_ocr_language_selection
-        self.skip_if_text_extracted = skip_if_text_extracted
-        if self.skip_if_text_extracted and isinstance(self.predictor, TextRecognizer):
-            raise ValueError(
-                "skip_if_text_extracted=True and TextRecognizer in TextExtractionService is not compatible"
-            )
 
     def serve(self, dp: Image) -> None:
         maybe_batched_text_rois = self.get_text_rois(dp)
@@ -154,11 +146,6 @@ class TextExtractionService(PipelineComponent):
         well `get_text_rois` will return an empty list.
         :return: list of ImageAnnotation or Image
         """
-        if self.skip_if_text_extracted:
-            text_categories = self.predictor.get_category_names()
-            text_anns = dp.get_annotation(category_names=text_categories)
-            if text_anns:
-                return []
 
         if self.extract_from_category:
             if self.predictor.accepts_batch:
@@ -223,7 +210,11 @@ class TextExtractionService(PipelineComponent):
         predictor = self.predictor.clone()
         if not isinstance(predictor, (ObjectDetector, PdfMiner, TextRecognizer)):
             raise ImageError(f"predictor must be of type ObjectDetector or PdfMiner, but is of type {type(predictor)}")
-        return self.__class__(predictor, deepcopy(self.extract_from_category), self.run_time_ocr_language_selection)
+        return self.__class__(
+            text_extract_detector=predictor,
+            extract_from_roi=deepcopy(self.extract_from_category),
+            run_time_ocr_language_selection=self.run_time_ocr_language_selection,
+        )
 
     def clear_predictor(self) -> None:
         self.predictor.clear_model()

deepdoctection/pipe/transform.py

@@ -1,5 +1,5 @@
 # -*- coding: utf-8 -*-
-# File: transform.py
+# File: test_transform.py
 
 # Copyright 2022 Dr. Janis Meyer. All rights reserved.
 #
@@ -22,6 +22,7 @@ on images (e.g. deskew, de-noising or more general GAN like operations.
 
 from __future__ import annotations
 
+from .. import DetectionResult
 from ..datapoint.image import Image
 from ..extern.base import ImageTransformer
 from .base import MetaAnnotation, PipelineComponent
@@ -49,25 +50,46 @@ class SimpleTransformService(PipelineComponent):
         super().__init__(self._get_name(transform_predictor.name), self.transform_predictor.model_id)
 
     def serve(self, dp: Image) -> None:
-        if dp.annotations:
-            raise RuntimeError(
-                "SimpleTransformService receives datapoints with ÌmageAnnotations. This violates the "
-                "pipeline building API but this can currently be catched only at runtime. "
-                "Please make sure that this component is the first one in the pipeline."
-            )
-
         if dp.image is not None:
             detection_result = self.transform_predictor.predict(dp.image)
-            transformed_image = self.transform_predictor.transform(dp.image, detection_result)
+            transformed_image = self.transform_predictor.transform_image(dp.image, detection_result)
             self.dp_manager.datapoint.clear_image(True)
             self.dp_manager.datapoint.image = transformed_image
-            self.dp_manager.set_summary_annotation(
-                summary_key=self.transform_predictor.get_category_names()[0],
-                summary_name=self.transform_predictor.get_category_names()[0],
-                summary_number=None,
-                summary_value=getattr(detection_result, self.transform_predictor.get_category_names()[0].value, None),
-                summary_score=detection_result.score,
-            )
+            for category in self.transform_predictor.get_category_names():
+                self.dp_manager.set_summary_annotation(
+                    summary_key=category,
+                    summary_name=category,
+                    summary_number=None,
+                    summary_value=getattr(detection_result, category.value, None),
+                    summary_score=detection_result.score,
+                )
+            detect_results = []
+            for ann in dp.get_annotation():
+                box = ann.get_bounding_box()
+                if not box.absolute_coords:
+                    box = box.transform(dp.width, dp.height)
+                detect_results.append(
+                    DetectionResult(
+                        box=box.to_list(mode="xyxy"),
+                        class_name=ann.category_name,  # type: ignore
+                        score=ann.score,
+                        class_id=ann.category_id,
+                        uuid=ann.annotation_id,
+                    )
+                )
+            output_detect_results = self.transform_predictor.transform_coords(detect_results)
+            for detect_result in output_detect_results:
+                ann = dp.get_annotation(annotation_ids=detect_result.uuid)[0]
+                transformed_ann_id = self.dp_manager.set_image_annotation(detect_result)
+                if transformed_ann_id is None:
+                    print("here")
+                transformed_ann = self.dp_manager.datapoint.get_annotation(annotation_ids=transformed_ann_id)[0]
+
+                for key, sub_ann in ann.sub_categories.items():
+                    transformed_ann.dump_sub_category(key, sub_ann)
+                if ann.image is not None:
+                    dp.image_ann_to_image(transformed_ann.annotation_id, ann.image.image is not None)
+                ann.deactivate()
 
     def clone(self) -> SimpleTransformService:
         return self.__class__(self.transform_predictor)

deepdoctection/train/hf_detr_train.py

@@ -272,6 +272,7 @@ def train_hf_detr(
         pretrained_model_name_or_path=path_config_json,
         num_labels=len(id2label),
     )
+    config.use_timm_backbone = True
 
     if path_weights != "":
         model = TableTransformerForObjectDetection.from_pretrained(

deepdoctection/utils/settings.py

@@ -67,6 +67,11 @@ class PageType(ObjectTypes):
     DOCUMENT_TYPE = "document_type"
     LANGUAGE = "language"
     ANGLE = "angle"
+    SIZE = "size"
+    PAD_TOP = "pad_top"
+    PAD_BOTTOM = "pad_bottom"
+    PAD_LEFT = "pad_left"
+    PAD_RIGHT = "pad_right"
 
 
 @object_types_registry.register("SummaryType")

deepdoctection/utils/transform.py

@@ -1,5 +1,5 @@
 # -*- coding: utf-8 -*-
-# File: transform.py
+# File: test_transform.py
 
 # Copyright 2022 Dr. Janis Meyer. All rights reserved.
 #
@@ -23,17 +23,50 @@ of coordinates. Most have the ideas have been taken from
 
 from __future__ import annotations
 
+import inspect
 from abc import ABC, abstractmethod
-from typing import Literal, Optional, Union
+from typing import Literal, Optional, Set, Union
 
 import numpy as np
 import numpy.typing as npt
 from numpy import float32
 
+from .settings import ObjectTypes, PageType
 from .types import PixelValues
 from .viz import viz_handler
 
-__all__ = ["ResizeTransform", "InferenceResize", "PadTransform", "normalize_image"]
+__all__ = [
+    "point4_to_box",
+    "box_to_point4",
+    "ResizeTransform",
+    "InferenceResize",
+    "PadTransform",
+    "normalize_image",
+    "pad_image",
+    "BaseTransform",
+    "RotationTransform",
+]
+
+
+def box_to_point4(boxes: npt.NDArray[np.float32]) -> npt.NDArray[np.float32]:
+    """
+    :param boxes: nx4
+    :return: (nx4)x2
+    """
+    box = boxes[:, [0, 1, 2, 3, 0, 3, 2, 1]]
+    box = box.reshape((-1, 2))
+    return box
+
+
+def point4_to_box(points: npt.NDArray[np.float32]) -> npt.NDArray[np.float32]:
+    """
+    :param points: (nx4)x2
+    :return: nx4 boxes (x1y1x2y2)
+    """
+    points = points.reshape((-1, 4, 2))
+    min_xy = points.min(axis=1)  # nx2
+    max_xy = points.max(axis=1)  # nx2
+    return np.concatenate((min_xy, max_xy), axis=1)
 
 
 class BaseTransform(ABC):
@@ -51,6 +84,33 @@ class BaseTransform(ABC):
         """The transformation that should be applied to the image"""
         raise NotImplementedError()
 
+    @abstractmethod
+    def apply_coords(self, coords: npt.NDArray[float32]) -> npt.NDArray[float32]:
+        """Transformation that should be applied to coordinates. Coords are supposed to to be passed as like
+
+        np.array([[ulx_0,uly_0,lrx_0,lry_0],[ulx_1,uly_1,lrx_1,lry_1],...])
+        """
+        raise NotImplementedError()
+
+    @abstractmethod
+    def inverse_apply_coords(self, coords: npt.NDArray[float32]) -> npt.NDArray[float32]:
+        """Inverse transformation going back from coordinates of transformed image to original image. Coords are
+        supposed to to be passed as like
+
+            np.array([[ulx_0,uly_0,lrx_0,lry_0],[ulx_1,uly_1,lrx_1,lry_1],...])
+        """
+        raise NotImplementedError()
+
+    @abstractmethod
+    def get_category_names(self) -> tuple[ObjectTypes, ...]:
+        """Get category names"""
+        raise NotImplementedError()
+
+    def get_init_args(self) -> Set[str]:
+        """Return the names of the arguments of the constructor."""
+        args = inspect.signature(self.__init__).parameters.keys()  # type: ignore
+        return {arg for arg in args if arg != "self"}
+
 
 class ResizeTransform(BaseTransform):
     """
@@ -93,6 +153,16 @@ class ResizeTransform(BaseTransform):
         coords[:, 1] = coords[:, 1] * (self.new_h * 1.0 / self.h)
         return coords
 
+    def inverse_apply_coords(self, coords: npt.NDArray[float32]) -> npt.NDArray[float32]:
+        """Inverse transformation going back from coordinates of resized image to original image"""
+        coords[:, 0] = coords[:, 0] * (self.w * 1.0 / self.new_w)
+        coords[:, 1] = coords[:, 1] * (self.h * 1.0 / self.new_h)
+        return coords
+
+    def get_category_names(self) -> tuple[ObjectTypes, ...]:
+        """Get category names"""
+        return (PageType.SIZE,)
+
 
 class InferenceResize:
     """
@@ -155,7 +225,7 @@ def pad_image(image: PixelValues, top: int, right: int, bottom: int, left: int)
     :param bottom: Bottom pixel value to pad
     :param left: Left pixel value to pad
     """
-    return np.pad(image, ((left, right), (top, bottom), (0, 0)), "constant", constant_values=255)
+    return np.pad(image, ((top, bottom), (left, right), (0, 0)), "constant", constant_values=255)
 
 
 class PadTransform(BaseTransform):
@@ -165,60 +235,125 @@ class PadTransform(BaseTransform):
 
     def __init__(
         self,
-        top: int,
-        right: int,
-        bottom: int,
-        left: int,
-        mode: Literal["xyxy", "xywh"] = "xyxy",
+        pad_top: int,
+        pad_right: int,
+        pad_bottom: int,
+        pad_left: int,
     ):
         """
-        :param top: padding top image side
-        :param right: padding right image side
-        :param bottom: padding bottom image side
-        :param left: padding left image side
-        :param mode: bounding box mode. Needed for transforming coordinates.
+        :param pad_top: padding top image side
+        :param pad_right: padding right image side
+        :param pad_bottom: padding bottom image side
+        :param pad_left: padding left image side
         """
-        self.top = top
-        self.right = right
-        self.bottom = bottom
-        self.left = left
+        self.pad_top = pad_top
+        self.pad_right = pad_right
+        self.pad_bottom = pad_bottom
+        self.pad_left = pad_left
         self.image_width: Optional[int] = None
         self.image_height: Optional[int] = None
-        self.mode = mode
 
     def apply_image(self, img: PixelValues) -> PixelValues:
         """Apply padding to image"""
         self.image_width = img.shape[1]
         self.image_height = img.shape[0]
-        return pad_image(img, self.top, self.right, self.bottom, self.left)
+        return pad_image(img, self.pad_top, self.pad_right, self.pad_bottom, self.pad_left)
 
     def apply_coords(self, coords: npt.NDArray[float32]) -> npt.NDArray[float32]:
         """Transformation that should be applied to coordinates"""
-        if self.mode == "xyxy":
-            coords[:, 0] = coords[:, 0] + self.left
-            coords[:, 1] = coords[:, 1] + self.top
-            coords[:, 2] = coords[:, 2] + self.left
-            coords[:, 3] = coords[:, 3] + self.top
-        else:
-            coords[:, 0] = coords[:, 0] + self.left
-            coords[:, 1] = coords[:, 1] + self.top
+        coords[:, 0] = coords[:, 0] + self.pad_left
+        coords[:, 1] = coords[:, 1] + self.pad_top
+        coords[:, 2] = coords[:, 2] + self.pad_left
+        coords[:, 3] = coords[:, 3] + self.pad_top
         return coords
 
     def inverse_apply_coords(self, coords: npt.NDArray[float32]) -> npt.NDArray[float32]:
         """Inverse transformation going back from coordinates of padded image to original image"""
         if self.image_height is None or self.image_width is None:
             raise ValueError("Initialize image_width and image_height first")
-
-        if self.mode == "xyxy":
-            coords[:, 0] = np.maximum(coords[:, 0] - self.left, np.zeros(coords[:, 0].shape))
-            coords[:, 1] = np.maximum(coords[:, 1] - self.top, np.zeros(coords[:, 1].shape))
-            coords[:, 2] = np.minimum(coords[:, 2] - self.left, np.ones(coords[:, 2].shape) * self.image_width)
-            coords[:, 3] = np.minimum(coords[:, 3] - self.top, np.ones(coords[:, 3].shape) * self.image_height)
-        else:
-            coords[:, 0] = np.maximum(coords[:, 0] - self.left, np.zeros(coords[:, 0].shape))
-            coords[:, 1] = np.maximum(coords[:, 1] - self.top, np.zeros(coords[:, 1].shape))
+        coords[:, 0] = np.maximum(coords[:, 0] - self.pad_left, np.zeros(coords[:, 0].shape))
+        coords[:, 1] = np.maximum(coords[:, 1] - self.pad_top, np.zeros(coords[:, 1].shape))
+        coords[:, 2] = np.minimum(coords[:, 2] - self.pad_left, np.ones(coords[:, 2].shape) * self.image_width)
+        coords[:, 3] = np.minimum(coords[:, 3] - self.pad_top, np.ones(coords[:, 3].shape) * self.image_height)
         return coords
 
     def clone(self) -> PadTransform:
         """clone"""
-        return self.__class__(self.top, self.right, self.bottom, self.left, self.mode)
+        return self.__class__(self.pad_top, self.pad_right, self.pad_bottom, self.pad_left)
+
+    def get_category_names(self) -> tuple[ObjectTypes, ...]:
+        """Get category names"""
+        return (
+            PageType.PAD_TOP,
+            PageType.PAD_RIGHT,
+            PageType.PAD_LEFT,
+            PageType.PAD_BOTTOM,
+        )
+
+
+class RotationTransform(BaseTransform):
+    """
+    A transform for rotating images by 90, 180, 270, or 360 degrees.
+    """
+
+    def __init__(self, angle: Literal[90, 180, 270, 360]):
+        """
+        :param angle: angle to rotate the image. Must be one of 90, 180, 270, or 360 degrees.
+        :param mode: coordinate format - "xyxy" (bounding box), "xywh" (width/height), "poly" (polygon points)
+        """
+        self.angle = angle
+        self.image_width: Optional[int] = None
+        self.image_height: Optional[int] = None
+
+    def apply_image(self, img: PixelValues) -> PixelValues:
+        """Apply rotation to image"""
+        self.image_width = img.shape[1]
+        self.image_height = img.shape[0]
+        return viz_handler.rotate_image(img, self.angle)
+
+    def apply_coords(self, coords: npt.NDArray[float32]) -> npt.NDArray[float32]:
+        """Transformation that should be applied to coordinates"""
+        if self.image_width is None or self.image_height is None:
+            raise ValueError("Initialize image_width and image_height first")
+
+        if self.angle == 90:
+            coords[:, [0, 1, 2, 3]] = coords[:, [1, 0, 3, 2]]
+            coords[:, [1, 3]] = self.image_width - coords[:, [1, 3]]
+            coords[:, [0, 1, 2, 3]] = coords[:, [0, 3, 2, 1]]
+        elif self.angle == 180:
+            coords[:, [0, 2]] = self.image_width - coords[:, [0, 2]]
+            coords[:, [1, 3]] = self.image_height - coords[:, [1, 3]]
+            coords[:, [0, 1, 2, 3]] = coords[:, [2, 3, 0, 1]]
+        elif self.angle == 270:
+            coords[:, [0, 1, 2, 3]] = coords[:, [1, 0, 3, 2]]
+            coords[:, [0, 2]] = self.image_height - coords[:, [0, 2]]
+            coords[:, [0, 1, 2, 3]] = coords[:, [2, 1, 0, 3]]
+
+        return coords
+
+    def inverse_apply_coords(self, coords: npt.NDArray[float32]) -> npt.NDArray[float32]:
+        """Inverse transformation going back from coordinates of rotated image to original image"""
+        if self.image_width is None or self.image_height is None:
+            raise ValueError("Initialize image_width and image_height first")
+
+        if self.angle == 90:
+            coords[:, [0, 1, 2, 3]] = coords[:, [1, 0, 3, 2]]
+            coords[:, [0, 2]] = self.image_width - coords[:, [0, 2]]
+            coords[:, [0, 1, 2, 3]] = coords[:, [2, 1, 0, 3]]
+        elif self.angle == 180:
+            coords[:, [0, 2]] = self.image_width - coords[:, [0, 2]]
+            coords[:, [1, 3]] = self.image_height - coords[:, [1, 3]]
+            coords[:, [0, 1, 2, 3]] = coords[:, [2, 3, 0, 1]]
+        elif self.angle == 270:
+            coords[:, [0, 1, 2, 3]] = coords[:, [1, 0, 3, 2]]
+            coords[:, [1, 3]] = self.image_height - coords[:, [1, 3]]
+            coords[:, [0, 1, 2, 3]] = coords[:, [0, 3, 2, 1]]
+        return coords
+
+    def clone(self) -> RotationTransform:
+        """clone"""
+        return self.__class__(self.angle)
+
+    def get_category_names(self) -> tuple[ObjectTypes, ...]:
+        """Get category names"""
+        return (PageType.ANGLE,)

{deepdoctection-0.39.7.dist-info → deepdoctection-0.41.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: deepdoctection
-Version: 0.39.7
+Version: 0.41.0
 Summary: Repository for Document AI
 Home-page: https://github.com/deepdoctection/deepdoctection
 Author: Dr. Janis Meyer

{deepdoctection-0.39.7.dist-info → deepdoctection-0.41.0.dist-info}/RECORD

@@ -1,9 +1,9 @@
-deepdoctection/__init__.py,sha256=SgzaP1SOePibE0bw0H_Jecy-ZmPWzhJYsZZ5UT_XLJs,12754
+deepdoctection/__init__.py,sha256=cr4wBN6EV5-nllpbHfU9jPaHa-L0QfPKcdqFThHeJFc,12850
 deepdoctection/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 deepdoctection/analyzer/__init__.py,sha256=icClxrd20XutD6LxLgEPIWceSs4j_QfI3szCE-9BL2w,729
-deepdoctection/analyzer/_config.py,sha256=1rfvVrp7cI2YLzpahD77aa1tZ_KFAIQ21DM1NWhxYiI,5058
+deepdoctection/analyzer/_config.py,sha256=kxQzDQvl2ygH84VTnumbRF7JLGM6VeJoBzv1xssm6H4,5019
 deepdoctection/analyzer/dd.py,sha256=bfR7e1JV7BwUNDRLu0jYZU7qQXnyA_vbRAJl2Ylrq5o,5905
-deepdoctection/analyzer/factory.py,sha256=7L-bJ9957TBn_C6OGWJFmZobrh8MPq4Q-Espx5faEiY,32435
+deepdoctection/analyzer/factory.py,sha256=aw6Nrqfjcg3M0Qy6214nX-oBv3GKTv62Wnz6WRTw2c8,33267
 deepdoctection/configs/__init__.py,sha256=TX_P6tqDOF1LK1mi9ruAl7x0mtv1Asm8cYWCz3Pe2dk,646
 deepdoctection/configs/conf_dd_one.yaml,sha256=qnrDAST1PHBtdIKE_hdkZexW22FqVvNTI-PEo9wvinM,3025
 deepdoctection/configs/conf_tesseract.yaml,sha256=oF6szDyoi15FHvq7yFUNIEjfA_jNLhGxoowiRsz_zY4,35
@@ -18,12 +18,12 @@ deepdoctection/dataflow/stats.py,sha256=Bsr6v7lcesKXUYtO9wjqlzx_Yq_uyIF3Lel-tQ0i
 deepdoctection/datapoint/__init__.py,sha256=3K406GbOPhoEp8koVaSbMocmSsmWifnSZ1SPb7C1lOY,1643
 deepdoctection/datapoint/annotation.py,sha256=FEgz4COxVDfjic0gG7kS6iHnWLBIgFnquQ63Cbj2a4Y,22531
 deepdoctection/datapoint/box.py,sha256=XPhC_xHqLZJjzafg1pIS_CxnVB5-0_yk-twsZZ3ncUU,30093
-deepdoctection/datapoint/convert.py,sha256=Be2FvmRXt-5prZ1vwa5fG6VjgEQ_31hiQ13hAoXoaes,7740
-deepdoctection/datapoint/image.py,sha256=uGmlgF6zGptvNowZTqf-io4hbd8aFFngAvQqgdEQ5Kw,34040
-deepdoctection/datapoint/view.py,sha256=sK6Ta9R6jdOS7iwF05-uPjL2wSz8wHQ5RIGCatw7i2M,50774
+deepdoctection/datapoint/convert.py,sha256=h3ky-Qn6YA8Qoyy5SMUkjJq___cK0hbcwFygDyqqm-4,7123
+deepdoctection/datapoint/image.py,sha256=_jN46UJUsOi6GC6VEUcp3L_vLL-iYRW05RKcFLWb6Dc,34048
+deepdoctection/datapoint/view.py,sha256=-4EygREbqQ0gNwS_LUWBX9svmug8wAE2vPA3k4zyyc4,50762
 deepdoctection/datasets/__init__.py,sha256=-A3aR90aDsHPmVM35JavfnQ2itYSCn3ujl4krRni1QU,1076
 deepdoctection/datasets/adapter.py,sha256=Ly_vbOAgVI73V41FUccnSX1ECTOyesW_qsuvQuvOZbw,7796
-deepdoctection/datasets/base.py,sha256=AZx-hw8Mchzb7FiOASt7zCbiybFNsM_diBzKXyC-auU,22618
+deepdoctection/datasets/base.py,sha256=X5Sr0yyal9x8rqWaWYr5mA_bE11UzN1iFXmWu605q2Y,22713
 deepdoctection/datasets/dataflow_builder.py,sha256=cYU2zV3gZW2bFvMHimlO9VIl3BAUaCwML08cCIQ8Em4,4107
 deepdoctection/datasets/info.py,sha256=sC1QCOdLWFMooVmiShZ43sLUpAi3FK4d0fsLyl_9-gA,20548
 deepdoctection/datasets/registry.py,sha256=utiB-PnE6vc5HvjcudO0O4Urp2BC3snqswY6d8uPQAo,3388
@@ -50,17 +50,17 @@ deepdoctection/eval/registry.py,sha256=v4mp-s67vBVRu1nQzuGlYPViQnMSeIXEcF_WmvfUC
 deepdoctection/eval/tedsmetric.py,sha256=rKw-734Y9CpBtIfkBSPQF2vAZxnIdWrI9Zc723P7RxI,9529
 deepdoctection/eval/tp_eval_callback.py,sha256=SXsXumoyxq-MIH9Cep5eUOwnNshMbKmC6mYOGwCg0pM,5283
 deepdoctection/extern/__init__.py,sha256=9Iks9b4Q_LynjcV167TVCoK8YsQRUcA2jjmAmDNA_X8,1056
-deepdoctection/extern/base.py,sha256=ONPgappl_P5HSwQr42FatuRnwMTvUPecPsCztDTN0Hw,24108
+deepdoctection/extern/base.py,sha256=oRuoAduVchsR3H7Ddm-KAe_smt0N6PlQftPqJ75FWfA,28944
 deepdoctection/extern/d2detect.py,sha256=zrKv1yurApnjD7QZIZk_8LYCahjmN82MQUjHjv8zvkQ,22127
-deepdoctection/extern/deskew.py,sha256=sPoixu8S9he-0wbs-jgxtPE2V9BiP4-3uZlb6F5Y1SA,3077
-deepdoctection/extern/doctrocr.py,sha256=R1PgKBFxVr_1-frkGvGL2ZBS19jpiktPQ4sJz_nBiNs,24622
+deepdoctection/extern/deskew.py,sha256=iax1ztkguGDfD4KQMDyuvanR4J2VgpCRuVWWDTwViu4,3083
+deepdoctection/extern/doctrocr.py,sha256=iUww7PuhCioEEv8ModrAElPPeEBa32zwUEa6OQf_y_c,24672
 deepdoctection/extern/fastlang.py,sha256=F4gK-SEwcCujjxH327ZDzMGWToJ49xS_dCKcePQ9IlY,4780
 deepdoctection/extern/hfdetr.py,sha256=JzHrrTyzS9qh6T2TsvKboAGZkIhno2txmSoLQ5Vd-lo,12077
 deepdoctection/extern/hflayoutlm.py,sha256=tFaf90FRbZzhSycdp8rGkeiPywQa6UcTEEwbayIXkr0,57023
 deepdoctection/extern/hflm.py,sha256=kwS6kcSlY_2m9u0RzBLTRq-UMM7c1PhyUaDTvSdejus,9217
 deepdoctection/extern/model.py,sha256=lbVwDa3vD6VwCD_dsozcI8b4xDZs4KJ1628SxaDdtHQ,55378
 deepdoctection/extern/pdftext.py,sha256=KS_t27SUiYn_IOS_J2lF9lSSo22vLagxmxvYCY3CqXA,7228
-deepdoctection/extern/tessocr.py,sha256=tG7etMvZ-jHFdq-jJAHYMJii3ujDjMfAFYUsjBp3nKI,17444
+deepdoctection/extern/tessocr.py,sha256=gRYJsk0jBRMG_ZLXbuJeRYPSPuVjXNwThs4op1hHpoA,17450
 deepdoctection/extern/texocr.py,sha256=yMt5ZzKtsjd7ogrcNXba7zccGGGF9LXK194EtER6YNQ,5804
 deepdoctection/extern/tpdetect.py,sha256=yAk1duQdoX-_pHLHgvhU7OOSiDy863q6XUMpjpYR734,8477
 deepdoctection/extern/pt/__init__.py,sha256=3Cu0ZHjbYsJomru7-RQXEHihEQLegZrmLetlHiqS58I,742
@@ -72,7 +72,7 @@ deepdoctection/extern/tp/tpcompat.py,sha256=rPW_JrYtz9PbV20dZiMKm6DTrjS1A3rAdhrh
 deepdoctection/extern/tp/tpfrcnn/__init__.py,sha256=OzDaR5A8HGz9a4VwjLiR9rN1Nf1cSebv8DVEMxStFOw,703
 deepdoctection/extern/tp/tpfrcnn/common.py,sha256=fCxwi2u752ZlI_DtIkLC_x9j9tyo1nnirAi2PmnziD4,3830
 deepdoctection/extern/tp/tpfrcnn/predict.py,sha256=957dnhCByS-FZH13efFWADhodaV4lKto-ikLPetfvEQ,4338
-deepdoctection/extern/tp/tpfrcnn/preproc.py,sha256=rMciBHGzNLRax8zpRrCv16lMnDHGimovIqLTp853WtA,12011
+deepdoctection/extern/tp/tpfrcnn/preproc.py,sha256=oHN9keBurjdNQqXmsb5BgURB5nl-eEp0KHvO1DPRQL4,12009
 deepdoctection/extern/tp/tpfrcnn/config/__init__.py,sha256=RhJiXId6vUSw_Pi49SPwj0jrf61VxxptXoGeBKtT42M,705
 deepdoctection/extern/tp/tpfrcnn/config/config.py,sha256=-T8AwNAIPR-_5OL1oEqm-Qe9GbN6JjAPVUuUw_XfMVc,11405
 deepdoctection/extern/tp/tpfrcnn/modeling/__init__.py,sha256=RhJiXId6vUSw_Pi49SPwj0jrf61VxxptXoGeBKtT42M,705
@@ -92,9 +92,9 @@ deepdoctection/mapper/cats.py,sha256=O06WGkpOIlSNMCy5VESl2HYOFDTuT9ls4aZIaWUv9VU
 deepdoctection/mapper/cocostruct.py,sha256=GcbUpPFUg67pcOHQluWBFOFcGaYnlZcTmwBDERBVgCA,5978
 deepdoctection/mapper/d2struct.py,sha256=Dx-YnycsIQH4a5-9Gn_yMhiQ-gOFgMueNeH3rhXjuCU,8555
 deepdoctection/mapper/hfstruct.py,sha256=2PjGKsYturVJBimLT1CahYh09KSRAFEHz_QNtC162kQ,5551
-deepdoctection/mapper/laylmstruct.py,sha256=abMZkYU2W0e_VcCm_c0ZXNFuv-lfMFWcTedcZS5EYvE,42935
+deepdoctection/mapper/laylmstruct.py,sha256=Es_aQOsfCkereJLOd1yaXhNAEEFJkODRuThUJ-d6hHU,42904
 deepdoctection/mapper/maputils.py,sha256=eI6ZcDg9W5uB6xQNBZpMIdEd86HlCxTtkJuyROdTqiw,8146
-deepdoctection/mapper/match.py,sha256=Ed9FsuVPNp_faaW5PKnvUHZoEXcRcrO-muduTMzjp1s,8937
+deepdoctection/mapper/match.py,sha256=TBc2yAWdQbM3sS64TerOJZwhPPMxk9cmfSXctKdwIU8,10269
 deepdoctection/mapper/misc.py,sha256=vX-fV420Te00eD-cqTiWBV2twHqdBcBV2_7rAFRgPRg,7164
 deepdoctection/mapper/pascalstruct.py,sha256=TzVU1p0oiw0nOuxTFFbEB9vXJxH1v6VUvTJ7MD0manU,3828
 deepdoctection/mapper/prodigystruct.py,sha256=Re4Sd_zAp6qOvbXZLmMJeG0IGEfMQxebuyDeZgMcTa8,6827
@@ -102,24 +102,24 @@ deepdoctection/mapper/pubstruct.py,sha256=PAJ2N1HSPNS6F2ZrIwlD7PiBhIM-rJscK_Ti8O
 deepdoctection/mapper/tpstruct.py,sha256=YNABRibvcISD5Lavg3jouoE4FMdqXEJoM-hNoB_rnww,4481
 deepdoctection/mapper/xfundstruct.py,sha256=_3r3c0K82fnF2h1HxA85h-9ETYrHwcERa6MNc6Ko6Z8,8807
 deepdoctection/pipe/__init__.py,sha256=ywTVoetftdL6plXg2YlBzMfmqBZupq7yXblSVyvvkcQ,1127
-deepdoctection/pipe/anngen.py,sha256=3319l4aaXzcY4w6ItVBNPX8LGS5fHFDVtyVY9KMefac,16393
+deepdoctection/pipe/anngen.py,sha256=7wvp7eghDwrgcIyu1vjRxmVy4SADPbn-k4ud8y2bgjU,15338
 deepdoctection/pipe/base.py,sha256=wlza9aDOKnHKrXmaz8MLyLz0nMqqcIWQ-6Lu944aicE,15390
-deepdoctection/pipe/common.py,sha256=C1KxEfJFSPeDqlnkiJ1ZYPuA36P8BU_4jVhdsszW_V8,17752
+deepdoctection/pipe/common.py,sha256=lY4kvQ5iOxp1NtdZf1KPrWXHMjU7Px8NtEBdUTDl9RQ,21032
 deepdoctection/pipe/concurrency.py,sha256=AAKRsVgaBEYNluntbDa46SBF1JZ_XqnWLDSWrNvAzEo,9657
 deepdoctection/pipe/doctectionpipe.py,sha256=bGW3ugky-fb-nEe-3bvO6Oc_4_6w82cQboGM_6p2eIo,12530
 deepdoctection/pipe/language.py,sha256=5zI0UQC6Fh12_r2pfVL42HoCGz2hpHrOhpXAn5m-rYw,5451
-deepdoctection/pipe/layout.py,sha256=xIhnJpyUSbvLbhTXyAKXY1hmG9352jihGYFSclTH_1g,5567
+deepdoctection/pipe/layout.py,sha256=ThULc0b1f9KyaXYk9z0qbuJ0nhIodah9PcrEq2xKpAY,5670
 deepdoctection/pipe/lm.py,sha256=x9NoYpivdjQF1r76a7PPrUuBEmuHP7ZukuXFDkXhXBc,17572
-deepdoctection/pipe/order.py,sha256=PnJZiCnxFluJiECXLTZT0c1Rr66vIRBFraa_G41UA2k,40121
+deepdoctection/pipe/order.py,sha256=0KNiMinedjfuDVVHxJSaDL1yl4Sub-miMPcEC4gGwPA,39423
 deepdoctection/pipe/refine.py,sha256=dTfI396xydPdbzpfo4yqFcuxl3UAB1y-WbSQn1o76ec,22367
 deepdoctection/pipe/registry.py,sha256=aFx-Tn0xhVA5l5H18duNW5QoTNKQltybsEUEzsMgUfg,902
-deepdoctection/pipe/segment.py,sha256=mWYRg7UR80PtIj1SIg_hiujDcCtLlvKJUP9vx4ZpW0Y,59318
-deepdoctection/pipe/sub_layout.py,sha256=ldFFuFIW5em2Rl1O1BYwclrXJ86wg-1RmDZmv35Cruw,13850
-deepdoctection/pipe/text.py,sha256=h9q6d3HFOs7LOg-iwdLUPiQxrPqgunBVNmtYMBrfRQE,11180
-deepdoctection/pipe/transform.py,sha256=9Om7X7hJeL4jgUwHM1CHa4sb5v7Qo1PtVG0ls_3nI7w,3798
+deepdoctection/pipe/segment.py,sha256=sny59GuP7dxLGX3YjHF0wllPxSiXL1GNQEhMGKcF8ZU,59594
+deepdoctection/pipe/sub_layout.py,sha256=D73H5b2Zl35fN58TaY0_nGhwI9Nwj3wqDdDPz8ce9Fg,13538
+deepdoctection/pipe/text.py,sha256=tLlJtneM__WsrAvp4pQFqwNlmq2RLqKqiPXlJ2lkniU,10483
+deepdoctection/pipe/transform.py,sha256=eCSRbyxHLz11owOHFA9UDX7tOJPZG2eiPWIGJv2odbk,4890
 deepdoctection/train/__init__.py,sha256=YFTRAZF1F7cEAKTdAIi1BLyYb6rSRcwq09Ui5Lu8d6E,1071
 deepdoctection/train/d2_frcnn_train.py,sha256=sFc_G-mEpaM8d1CCE0_6Gl4nBh11X2RYRBA3p_ylFJQ,16000
-deepdoctection/train/hf_detr_train.py,sha256=NEOoRjZ00bPwN1supTJD7VIcHRgvDJFSYcugiHo_Rqs,12007
+deepdoctection/train/hf_detr_train.py,sha256=uBkkRyxrJF5UF__KbYvIlmb-HRWQ9TY6LiJr1Rm56kI,12043
 deepdoctection/train/hf_layoutlm_train.py,sha256=8kiGp_8GEyqCkLgeMgCJOLJWSVoKWkUBHsZtDjZOcRk,22556
 deepdoctection/train/tp_frcnn_train.py,sha256=pEpXokSVGveqo82pRnhnAmHPmjQ_8wQWpqM4ZyNHJgs,13049
 deepdoctection/utils/__init__.py,sha256=brBceRWeov9WXMiJTjyJOF2rHMP8trGGRRjhMdZ61nI,2371
@@ -135,14 +135,14 @@ deepdoctection/utils/logger.py,sha256=J0OVKiXP_2A82MWbbJoOeMEJ-75aZu5npgaS_yI6mV
 deepdoctection/utils/metacfg.py,sha256=hD76KQ_RnD_5B02qLI2Zxf3WfnsnXhEI_KUTKpw91RI,5711
 deepdoctection/utils/mocks.py,sha256=IkN3-IzAl4eX0ibgKIHg8IY7ykVw6BnpF6XnxKnKaZI,2389
 deepdoctection/utils/pdf_utils.py,sha256=Fi0eZ2GbnO7N61Rd8b8YRKRff4dalHAzkcn3zpGPoic,13119
-deepdoctection/utils/settings.py,sha256=hDD6yDX_4pQXwR5ILVwJIj6hb7NXA0-ifnC25ldcUjA,12464
+deepdoctection/utils/settings.py,sha256=nmA7aqzHTMrAhodNbPjr-afLRt2EWjpx8ipQeE-bqqs,12590
 deepdoctection/utils/tqdm.py,sha256=cBUtR0L1x0KMeYrLP2rrzyzCamCjpQAKroHXLv81_pk,1820
-deepdoctection/utils/transform.py,sha256=3kCgsEeRkG1efCdkfvj7tUFMs-e2jbjbflq826F2GPU,8502
+deepdoctection/utils/transform.py,sha256=5mY5D6hhk6cKFp0T1LJ2_jMjjBxJopcFZffAN5PKvFU,13699
 deepdoctection/utils/types.py,sha256=ti4WdtIJSg3TGK_YPkkoY9PYGMnR2tTX6Xfik8U1pNk,2986
 deepdoctection/utils/utils.py,sha256=csVs_VvCq4QBETPoE2JdTTL4MFYnD4xh-Js5vRb612g,6492
 deepdoctection/utils/viz.py,sha256=Jf8ePNYWlpuyaS6SeTYQ4OyA3eNhtgjvAQZnGNdgHC0,27051
-deepdoctection-0.39.7.dist-info/licenses/LICENSE,sha256=GQ0rUvuGdrMNEI3iHK5UQx6dIMU1QwAuyXsxUHn5MEQ,11351
-deepdoctection-0.39.7.dist-info/METADATA,sha256=f1bypRgjWbclKDMdqJMW_CwoIOCnE88r_C6PA9qGvDY,19763
-deepdoctection-0.39.7.dist-info/WHEEL,sha256=DK49LOLCYiurdXXOXwGJm6U4DkHkg4lcxjhqwRa0CP4,91
-deepdoctection-0.39.7.dist-info/top_level.txt,sha256=hs2DdoOL9h4mnHhmO82BT4pz4QATIoOZ20PZmlnxFI8,15
-deepdoctection-0.39.7.dist-info/RECORD,,
+deepdoctection-0.41.0.dist-info/licenses/LICENSE,sha256=GQ0rUvuGdrMNEI3iHK5UQx6dIMU1QwAuyXsxUHn5MEQ,11351
+deepdoctection-0.41.0.dist-info/METADATA,sha256=-okfYzsN5uUB9BR6j3eKaLZtzxCr9lRWsoyvf-RgXwM,19763
+deepdoctection-0.41.0.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
+deepdoctection-0.41.0.dist-info/top_level.txt,sha256=hs2DdoOL9h4mnHhmO82BT4pz4QATIoOZ20PZmlnxFI8,15
+deepdoctection-0.41.0.dist-info/RECORD,,

{deepdoctection-0.39.7.dist-info → deepdoctection-0.41.0.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (78.0.2)
+Generator: setuptools (78.1.0)
 Root-Is-Purelib: true
 Tag: py3-none-any