deepdoctection 0.35__py3-none-any.whl → 0.36__py3-none-any.whl

deepdoctection/__init__.py

@@ -24,14 +24,10 @@ from .utils.logger import LoggingRecord, logger
 
 # pylint: enable=wrong-import-position
 
-__version__ = 0.35
+__version__ = 0.36
 
 _IMPORT_STRUCTURE = {
-    "analyzer": [
-        "config_sanity_checks",
-        "get_dd_analyzer",
-        "ServiceFactory"
-    ],
+    "analyzer": ["config_sanity_checks", "get_dd_analyzer", "ServiceFactory"],
     "configs": [],
     "dataflow": [
         "DataFlowTerminated",
@@ -379,6 +375,7 @@ _IMPORT_STRUCTURE = {
         "get_pdf_file_writer",
         "PDFStreamer",
         "pdf_to_np_array",
+        "split_pdf",
         "ObjectTypes",
         "TypeOrStr",
         "object_types_registry",

deepdoctection/analyzer/_config.py

@@ -17,11 +17,13 @@
 
 """Pipeline configuration for deepdoctection analyzer. Do not change the defaults in this file. """
 
+from ..datapoint.view import IMAGE_DEFAULTS
 from ..utils.metacfg import AttrDict
 from ..utils.settings import CellType, LayoutType
 
 cfg = AttrDict()
 
+
 cfg.LANGUAGE = None
 cfg.LIB = None
 cfg.DEVICE = None
@@ -32,11 +34,9 @@ cfg.USE_TABLE_SEGMENTATION = True
 cfg.TF.LAYOUT.WEIGHTS = "layout/model-800000_inf_only.data-00000-of-00001"
 cfg.TF.LAYOUT.FILTER = None
 
-
 cfg.TF.CELL.WEIGHTS = "cell/model-1800000_inf_only.data-00000-of-00001"
 cfg.TF.CELL.FILTER = None
 
-
 cfg.TF.ITEM.WEIGHTS = "item/model-1620000_inf_only.data-00000-of-00001"
 cfg.TF.ITEM.FILTER = None
 
@@ -112,7 +112,7 @@ cfg.OCR.WEIGHTS.DOCTR_WORD.PT = "doctr/db_resnet50/pt/db_resnet50-ac60cadc.pt"
 cfg.OCR.WEIGHTS.DOCTR_RECOGNITION.TF = "doctr/crnn_vgg16_bn/tf/crnn_vgg16_bn-76b7f2c6.zip"
 cfg.OCR.WEIGHTS.DOCTR_RECOGNITION.PT = "doctr/crnn_vgg16_bn/pt/crnn_vgg16_bn-9762b0b0.pt"
 
-cfg.TEXT_CONTAINER = LayoutType.WORD
+cfg.TEXT_CONTAINER = IMAGE_DEFAULTS["text_container"]
 cfg.WORD_MATCHING.PARENTAL_CATEGORIES = [
     LayoutType.TEXT,
     LayoutType.TITLE,
@@ -127,24 +127,16 @@ cfg.WORD_MATCHING.RULE = "ioa"
 cfg.WORD_MATCHING.THRESHOLD = 0.6
 cfg.WORD_MATCHING.MAX_PARENT_ONLY = True
 
-cfg.TEXT_ORDERING.TEXT_BLOCK_CATEGORIES = [
-    LayoutType.TEXT,
-    LayoutType.TITLE,
-    LayoutType.LIST,
-    LayoutType.CELL,
-    CellType.COLUMN_HEADER,
-    CellType.PROJECTED_ROW_HEADER,
-    CellType.SPANNING,
-    CellType.ROW_HEADER,
-]
-cfg.TEXT_ORDERING.FLOATING_TEXT_BLOCK_CATEGORIES = [
-    LayoutType.TEXT,
-    LayoutType.TITLE,
-    LayoutType.LIST,
-]
+cfg.TEXT_ORDERING.TEXT_BLOCK_CATEGORIES = IMAGE_DEFAULTS["text_block_categories"]
+cfg.TEXT_ORDERING.FLOATING_TEXT_BLOCK_CATEGORIES = IMAGE_DEFAULTS["floating_text_block_categories"]
 cfg.TEXT_ORDERING.INCLUDE_RESIDUAL_TEXT_CONTAINER = False
 cfg.TEXT_ORDERING.STARTING_POINT_TOLERANCE = 0.005
 cfg.TEXT_ORDERING.BROKEN_LINE_TOLERANCE = 0.003
 cfg.TEXT_ORDERING.HEIGHT_TOLERANCE = 2.0
 cfg.TEXT_ORDERING.PARAGRAPH_BREAK = 0.035
+
+cfg.USE_LAYOUT_LINK = False
+cfg.LAYOUT_LINK.PARENTAL_CATEGORIES = []
+cfg.LAYOUT_LINK.CHILD_CATEGORIES = []
+
 cfg.freeze()

deepdoctection/analyzer/factory.py

@@ -33,7 +33,13 @@ from ..extern.tessocr import TesseractOcrDetector, TesseractRotationTransformer
 from ..extern.texocr import TextractOcrDetector
 from ..extern.tpdetect import TPFrcnnDetector
 from ..pipe.base import PipelineComponent
-from ..pipe.common import AnnotationNmsService, IntersectionMatcher, MatchingService, PageParsingService
+from ..pipe.common import (
+    AnnotationNmsService,
+    IntersectionMatcher,
+    MatchingService,
+    NeighbourMatcher,
+    PageParsingService,
+)
 from ..pipe.doctectionpipe import DoctectionPipe
 from ..pipe.layout import ImageLayoutService
 from ..pipe.order import TextOrderService
@@ -78,7 +84,7 @@ class ServiceFactory:
     """
 
     @staticmethod
-    def build_layout_detector(
+    def _build_layout_detector(
         config: AttrDict,
         mode: str,
     ) -> Union[D2FrcnnDetector, TPFrcnnDetector, HFDetrDerivedDetector, D2FrcnnTracingDetector]:
@@ -141,17 +147,38 @@ class ServiceFactory:
         )
 
     @staticmethod
-    def build_rotation_detector() -> TesseractRotationTransformer:
+    def build_layout_detector(
+        config: AttrDict, mode: str
+    ) -> Union[D2FrcnnDetector, TPFrcnnDetector, HFDetrDerivedDetector, D2FrcnnTracingDetector]:
+        """Building a layout detector according to the config
+
+        :param config: configuration object
+        :param mode: either `LAYOUT`,`CELL` or `ITEM`
+        """
+        return ServiceFactory._build_layout_detector(config, mode)
+
+    @staticmethod
+    def _build_rotation_detector() -> TesseractRotationTransformer:
         """Building a rotation detector"""
         return TesseractRotationTransformer()
 
     @staticmethod
-    def build_transform_service(transform_predictor: ImageTransformer) -> SimpleTransformService:
+    def build_rotation_detector() -> TesseractRotationTransformer:
+        """Building a rotation detector"""
+        return ServiceFactory._build_rotation_detector()
+
+    @staticmethod
+    def _build_transform_service(transform_predictor: ImageTransformer) -> SimpleTransformService:
         """Building a transform service with a given predictor"""
         return SimpleTransformService(transform_predictor)
 
     @staticmethod
-    def build_padder(config: AttrDict, mode: str) -> PadTransform:
+    def build_transform_service(transform_predictor: ImageTransformer) -> SimpleTransformService:
+        """Building a transform service with a given predictor"""
+        return ServiceFactory._build_transform_service(transform_predictor)
+
+    @staticmethod
+    def _build_padder(config: AttrDict, mode: str) -> PadTransform:
         """Building a padder according to the config
 
         :param config: configuration object
@@ -164,10 +191,20 @@ class ServiceFactory:
             getattr(config.PT, mode).PAD.BOTTOM,
             getattr(config.PT, mode).PAD.LEFT,
         )
-        return PadTransform(top=top, right=right, bottom=bottom, left=left)
+        return PadTransform(top=top, right=right, bottom=bottom, left=left)  #
 
     @staticmethod
-    def build_layout_service(config: AttrDict, detector: ObjectDetector, mode: str) -> ImageLayoutService:
+    def build_padder(config: AttrDict, mode: str) -> PadTransform:
+        """Building a padder according to the config
+
+        :param config: configuration object
+        :param mode: either `LAYOUT`,`CELL` or `ITEM`
+        :return `PadTransform` instance
+        """
+        return ServiceFactory._build_padder(config, mode)
+
+    @staticmethod
+    def _build_layout_service(config: AttrDict, detector: ObjectDetector, mode: str) -> ImageLayoutService:
         """Building a layout service with a given detector
 
         :param config: configuration object
@@ -181,7 +218,18 @@ class ServiceFactory:
         return ImageLayoutService(layout_detector=detector, to_image=True, crop_image=True, padder=padder)
 
     @staticmethod
-    def build_layout_nms_service(config: AttrDict) -> AnnotationNmsService:
+    def build_layout_service(config: AttrDict, detector: ObjectDetector, mode: str) -> ImageLayoutService:
+        """Building a layout service with a given detector
+
+        :param config: configuration object
+        :param detector: will be passed to the `ImageLayoutService`
+        :param mode: either `LAYOUT`,`CELL` or `ITEM`
+        :return `ImageLayoutService` instance
+        """
+        return ServiceFactory._build_layout_service(config, detector, mode)
+
+    @staticmethod
+    def _build_layout_nms_service(config: AttrDict) -> AnnotationNmsService:
         """Building a NMS service for layout annotations
 
         :param config: configuration object
@@ -199,7 +247,15 @@ class ServiceFactory:
         )
 
     @staticmethod
-    def build_sub_image_service(config: AttrDict, detector: ObjectDetector, mode: str) -> SubImageLayoutService:
+    def build_layout_nms_service(config: AttrDict) -> AnnotationNmsService:
+        """Building a NMS service for layout annotations
+
+        :param config: configuration object
+        """
+        return ServiceFactory._build_layout_nms_service(config)
+
+    @staticmethod
+    def _build_sub_image_service(config: AttrDict, detector: ObjectDetector, mode: str) -> SubImageLayoutService:
         """
         Building a sub image layout service with a given detector
 
@@ -226,7 +282,19 @@ class ServiceFactory:
         )
 
     @staticmethod
-    def build_ocr_detector(config: AttrDict) -> Union[TesseractOcrDetector, DoctrTextRecognizer, TextractOcrDetector]:
+    def build_sub_image_service(config: AttrDict, detector: ObjectDetector, mode: str) -> SubImageLayoutService:
+        """
+        Building a sub image layout service with a given detector
+
+        :param config: configuration object
+        :param detector: will be passed to the `SubImageLayoutService`
+        :param mode: either `LAYOUT`,`CELL` or `ITEM`
+        :return: `SubImageLayoutService` instance
+        """
+        return ServiceFactory._build_sub_image_service(config, detector, mode)
+
+    @staticmethod
+    def _build_ocr_detector(config: AttrDict) -> Union[TesseractOcrDetector, DoctrTextRecognizer, TextractOcrDetector]:
         """
         Building OCR predictor
 
@@ -266,6 +334,15 @@ class ServiceFactory:
             return TextractOcrDetector(**credentials_kwargs)
         raise ValueError("You have set USE_OCR=True but any of USE_TESSERACT, USE_DOCTR, USE_TEXTRACT is set to False")
 
+    @staticmethod
+    def build_ocr_detector(config: AttrDict) -> Union[TesseractOcrDetector, DoctrTextRecognizer, TextractOcrDetector]:
+        """
+        Building OCR predictor
+
+        :param config: configuration object
+        """
+        return ServiceFactory._build_ocr_detector(config)
+
     @staticmethod
     def build_doctr_word_detector(config: AttrDict) -> DoctrTextlineDetector:
         """Building `DoctrTextlineDetector` instance
@@ -285,7 +362,7 @@ class ServiceFactory:
         )
 
     @staticmethod
-    def build_table_segmentation_service(
+    def _build_table_segmentation_service(
         config: AttrDict,
         detector: ObjectDetector,
     ) -> Union[PubtablesSegmentationService, TableSegmentationService]:
@@ -342,7 +419,32 @@ class ServiceFactory:
         return table_segmentation
 
     @staticmethod
-    def build_table_refinement_service(config: AttrDict) -> TableSegmentationRefinementService:
+    def build_table_segmentation_service(
+        config: AttrDict,
+        detector: ObjectDetector,
+    ) -> Union[PubtablesSegmentationService, TableSegmentationService]:
+        """
+        Build and return a table segmentation service based on the provided detector.
+
+        Depending on the type of the detector, this method will return either a `PubtablesSegmentationService` or a
+        `TableSegmentationService` instance. The selection is made as follows:
+
+        - If the detector is an instance of `HFDetrDerivedDetector`, a `PubtablesSegmentationService` is created and
+          returned. This service uses specific configuration parameters for segmentation, such as assignment rules,
+          thresholds, and cell names defined in the `cfg` object.
+        - For other detector types, a `TableSegmentationService` is created and returned. This service also uses
+          configuration parameters from the `cfg` object but is tailored for different segmentation needs.
+
+        :param config: configuration object
+        :param detector: An instance of `ObjectDetector` used to determine the type of table segmentation
+        service to build.
+        :return: An instance of either `PubtablesSegmentationService` or `TableSegmentationService` based on the
+                 detector type.
+        """
+        return ServiceFactory._build_table_segmentation_service(config, detector)
+
+    @staticmethod
+    def _build_table_refinement_service(config: AttrDict) -> TableSegmentationRefinementService:
         """Building a table segmentation refinement service
 
         :param config: configuration object
@@ -354,7 +456,16 @@ class ServiceFactory:
         )
 
     @staticmethod
-    def build_pdf_text_detector(config: AttrDict) -> PdfPlumberTextDetector:
+    def build_table_refinement_service(config: AttrDict) -> TableSegmentationRefinementService:
+        """Building a table segmentation refinement service
+
+        :param config: configuration object
+        :return: TableSegmentationRefinementService
+        """
+        return ServiceFactory._build_table_refinement_service(config)
+
+    @staticmethod
+    def _build_pdf_text_detector(config: AttrDict) -> PdfPlumberTextDetector:
         """Building a PDF text detector
 
         :param config: configuration object
@@ -365,7 +476,16 @@ class ServiceFactory:
         )
 
     @staticmethod
-    def build_pdf_miner_text_service(detector: PdfMiner) -> TextExtractionService:
+    def build_pdf_text_detector(config: AttrDict) -> PdfPlumberTextDetector:
+        """Building a PDF text detector
+
+        :param config: configuration object
+        :return: PdfPlumberTextDetector
+        """
+        return ServiceFactory._build_pdf_text_detector(config)
+
+    @staticmethod
+    def _build_pdf_miner_text_service(detector: PdfMiner) -> TextExtractionService:
         """Building a PDFMiner text extraction service
 
         :param detector: PdfMiner
@@ -373,6 +493,15 @@ class ServiceFactory:
         """
         return TextExtractionService(detector)
 
+    @staticmethod
+    def build_pdf_miner_text_service(detector: PdfMiner) -> TextExtractionService:
+        """Building a PDFMiner text extraction service
+
+        :param detector: PdfMiner
+        :return: TextExtractionService
+        """
+        return ServiceFactory._build_pdf_miner_text_service(detector)
+
     @staticmethod
     def build_doctr_word_detector_service(detector: DoctrTextlineDetector) -> ImageLayoutService:
         """Building a Doctr word detector service
@@ -385,7 +514,7 @@ class ServiceFactory:
         )
 
     @staticmethod
-    def build_text_extraction_service(
+    def _build_text_extraction_service(
         config: AttrDict, detector: Union[TesseractOcrDetector, DoctrTextRecognizer, TextractOcrDetector]
     ) -> TextExtractionService:
         """Building a text extraction service
@@ -401,7 +530,19 @@ class ServiceFactory:
         )
 
     @staticmethod
-    def build_word_matching_service(config: AttrDict) -> MatchingService:
+    def build_text_extraction_service(
+        config: AttrDict, detector: Union[TesseractOcrDetector, DoctrTextRecognizer, TextractOcrDetector]
+    ) -> TextExtractionService:
+        """Building a text extraction service
+
+        :param config: configuration object
+        :param detector: OCR detector
+        :return: TextExtractionService
+        """
+        return ServiceFactory._build_text_extraction_service(config, detector)
+
+    @staticmethod
+    def _build_word_matching_service(config: AttrDict) -> MatchingService:
         """Building a word matching service
 
         :param config: configuration object
@@ -420,7 +561,40 @@ class ServiceFactory:
         )
 
     @staticmethod
-    def build_text_order_service(config: AttrDict) -> TextOrderService:
+    def build_word_matching_service(config: AttrDict) -> MatchingService:
+        """Building a word matching service
+
+        :param config: configuration object
+        :return: MatchingService
+        """
+        return ServiceFactory._build_word_matching_service(config)
+
+    @staticmethod
+    def _build_layout_link_matching_service(config: AttrDict) -> MatchingService:
+        """Building a word matching service
+
+        :param config: configuration object
+        :return: MatchingService
+        """
+        neighbor_matcher = NeighbourMatcher()
+        return MatchingService(
+            parent_categories=config.LAYOUT_LINK.PARENTAL_CATEGORIES,
+            child_categories=config.LAYOUT_LINK.CHILD_CATEGORIES,
+            matcher=neighbor_matcher,
+            relationship_key=Relationships.LAYOUT_LINK,
+        )
+
+    @staticmethod
+    def build_layout_link_matching_service(config: AttrDict) -> MatchingService:
+        """Building a word matching service
+
+        :param config: configuration object
+        :return: MatchingService
+        """
+        return ServiceFactory._build_layout_link_matching_service(config)
+
+    @staticmethod
+    def _build_text_order_service(config: AttrDict) -> TextOrderService:
         """Building a text order service
 
         :param config: configuration object
@@ -438,7 +612,16 @@ class ServiceFactory:
         )
 
     @staticmethod
-    def build_page_parsing_service(config: AttrDict) -> PageParsingService:
+    def build_text_order_service(config: AttrDict) -> TextOrderService:
+        """Building a text order service
+
+        :param config: configuration object
+        :return: TextOrderService instance
+        """
+        return ServiceFactory._build_text_order_service(config)
+
+    @staticmethod
+    def _build_page_parsing_service(config: AttrDict) -> PageParsingService:
         """Building a page parsing service
 
         :param config: configuration object
@@ -450,6 +633,15 @@ class ServiceFactory:
             include_residual_text_container=config.TEXT_ORDERING.INCLUDE_RESIDUAL_TEXT_CONTAINER,
         )
 
+    @staticmethod
+    def build_page_parsing_service(config: AttrDict) -> PageParsingService:
+        """Building a page parsing service
+
+        :param config: configuration object
+        :return: PageParsingService instance
+        """
+        return ServiceFactory._build_page_parsing_service(config)
+
     @staticmethod
     def build_analyzer(config: AttrDict) -> DoctectionPipe:
         """
@@ -517,6 +709,10 @@ class ServiceFactory:
             text_order_service = ServiceFactory.build_text_order_service(config)
             pipe_component_list.append(text_order_service)
 
+        if config.USE_LAYOUT_LINK:
+            layout_link_matching_service = ServiceFactory.build_layout_link_matching_service(config)
+            pipe_component_list.append(layout_link_matching_service)
+
         page_parsing_service = ServiceFactory.build_page_parsing_service(config)
 
         return DoctectionPipe(pipeline_component_list=pipe_component_list, page_parsing_service=page_parsing_service)

deepdoctection/configs/conf_dd_one.yaml

@@ -98,3 +98,7 @@ TEXT_ORDERING:
   BROKEN_LINE_TOLERANCE: 0.003
   HEIGHT_TOLERANCE: 2.0
   PARAGRAPH_BREAK: 0.035
+  USE_LAYOUT_LINK: False
+  LAYOUT_LINK:
+    PARENTAL_CATEGORIES:
+    CHILD_CATEGORIES:

deepdoctection/datapoint/view.py

@@ -25,6 +25,7 @@ from copy import copy
 from typing import Any, Mapping, Optional, Sequence, Type, TypedDict, Union, no_type_check
 
 import numpy as np
+from typing_extensions import LiteralString
 
 from ..utils.error import AnnotationError, ImageError
 from ..utils.logger import LoggingRecord, logger
@@ -40,10 +41,12 @@ from ..utils.settings import (
     WordType,
     get_type,
 )
+from ..utils.transform import ResizeTransform
 from ..utils.types import HTML, AnnotationDict, Chunks, ImageDict, PathLikeOrStr, PixelValues, Text_, csv
 from ..utils.viz import draw_boxes, interactive_imshow, viz_handler
 from .annotation import CategoryAnnotation, ContainerAnnotation, ImageAnnotation, ann_from_dict
 from .box import BoundingBox, crop_box_from_image
+from .convert import box_to_point4, point4_to_box
 from .image import Image
 
 
@@ -101,7 +104,7 @@ class ImageAnnotationBaseView(ImageAnnotation):
             return np_image
         raise AnnotationError(f"base_page.image is None for {self.annotation_id}")
 
-    def __getattr__(self, item: str) -> Optional[Union[str, int, list[str]]]:
+    def __getattr__(self, item: str) -> Optional[Union[str, int, list[str], list[ImageAnnotationBaseView]]]:
         """
         Get attributes defined by registered `self.get_attribute_names()` in a multi step process:
 
@@ -126,6 +129,9 @@ class ImageAnnotationBaseView(ImageAnnotation):
             if isinstance(sub_cat, ContainerAnnotation):
                 return sub_cat.value
             return sub_cat.category_id
+        if item in self.relationships:
+            relationship_ids = self.get_relationship(get_type(item))
+            return self.base_page.get_annotation(annotation_ids=relationship_ids)
         if self.image is not None:
             if item in self.image.summary.sub_categories:
                 sub_cat = self.get_summary(get_type(item))
@@ -165,7 +171,11 @@ class Word(ImageAnnotationBaseView):
     """
 
     def get_attribute_names(self) -> set[str]:
-        return set(WordType).union(super().get_attribute_names()).union({Relationships.READING_ORDER})
+        return (
+            set(WordType)
+            .union(super().get_attribute_names())
+            .union({Relationships.READING_ORDER, Relationships.LAYOUT_LINK})
+        )
 
 
 class Layout(ImageAnnotationBaseView):
@@ -246,7 +256,11 @@ class Layout(ImageAnnotationBaseView):
         }
 
     def get_attribute_names(self) -> set[str]:
-        return {"words", "text"}.union(super().get_attribute_names()).union({Relationships.READING_ORDER})
+        return (
+            {"words", "text"}
+            .union(super().get_attribute_names())
+            .union({Relationships.READING_ORDER, Relationships.LAYOUT_LINK})
+        )
 
     def __len__(self) -> int:
         """len of text counted by number of characters"""
@@ -433,8 +447,8 @@ class ImageDefaults(TypedDict):
     """ImageDefaults"""
 
     text_container: LayoutType
-    floating_text_block_categories: tuple[LayoutType, ...]
-    text_block_categories: tuple[LayoutType, ...]
+    floating_text_block_categories: tuple[Union[LayoutType, CellType], ...]
+    text_block_categories: tuple[Union[LayoutType, CellType], ...]
 
 
 IMAGE_DEFAULTS: ImageDefaults = {
@@ -448,9 +462,13 @@ IMAGE_DEFAULTS: ImageDefaults = {
     "text_block_categories": (
         LayoutType.TEXT,
         LayoutType.TITLE,
-        LayoutType.FIGURE,
         LayoutType.LIST,
         LayoutType.CELL,
+        LayoutType.FIGURE,
+        CellType.COLUMN_HEADER,
+        CellType.PROJECTED_ROW_HEADER,
+        CellType.SPANNING,
+        CellType.ROW_HEADER,
     ),
 }
 
@@ -510,6 +528,8 @@ class Page(Image):
         "document_id",
         "page_number",
         "angle",
+        "figures",
+        "residual_layouts",
     }
     include_residual_text_container: bool = True
 
@@ -608,6 +628,41 @@ class Page(Image):
         """
         return self.get_annotation(category_names=LayoutType.TABLE)
 
+    @property
+    def figures(self) -> list[ImageAnnotationBaseView]:
+        """
+        A list of a figures.
+        """
+        return self.get_annotation(category_names=LayoutType.FIGURE)
+
+    @property
+    def residual_layouts(self) -> list[ImageAnnotationBaseView]:
+        """
+        A list of all residual layouts. Residual layouts are all layouts that are
+           - not floating text blocks,
+           - not text containers,
+           - not tables,
+           - not figures
+           - not cells
+           - not rows
+           - not columns
+        """
+        return self.get_annotation(category_names=self._get_residual_layout())
+
+    def _get_residual_layout(self) -> list[LiteralString]:
+        layouts = copy(list(self.floating_text_block_categories))
+        layouts.extend(
+            [
+                LayoutType.TABLE,
+                LayoutType.FIGURE,
+                self.text_container,
+                LayoutType.CELL,
+                LayoutType.ROW,
+                LayoutType.COLUMN,
+            ]
+        )
+        return [layout for layout in LayoutType if layout not in layouts]
+
     @classmethod
     def from_image(
         cls,
@@ -801,12 +856,15 @@ class Page(Image):
         self,
         show_tables: bool = True,
         show_layouts: bool = True,
+        show_figures: bool = False,
+        show_residual_layouts: bool = False,
         show_cells: bool = True,
         show_table_structure: bool = True,
         show_words: bool = False,
         show_token_class: bool = True,
         ignore_default_token_class: bool = False,
         interactive: bool = False,
+        scaled_width: int = 600,
         **debug_kwargs: str,
     ) -> Optional[PixelValues]:
         """
@@ -827,12 +885,14 @@ class Page(Image):
 
         :param show_tables: Will display all tables boxes as well as cells, rows and columns
         :param show_layouts: Will display all other layout components.
+        :param show_figures: Will display all figures
         :param show_cells: Will display cells within tables. (Only available if `show_tables=True`)
         :param show_table_structure: Will display rows and columns
         :param show_words: Will display bounding boxes around words labeled with token class and bio tag (experimental)
         :param show_token_class: Will display token class instead of token tags (i.e. token classes with tags)
         :param interactive: If set to True will open an interactive image, otherwise it will return a numpy array that
                             can be displayed differently.
+        :param scaled_width: Width of the image to display
         :param ignore_default_token_class: Will ignore displaying word bounding boxes with default or None token class
                                            label
         :return: If `interactive=False` will return a numpy array.
@@ -858,6 +918,11 @@ class Page(Image):
                 box_stack.append(item.bbox)
                 category_names_list.append(item.category_name.value)
 
+        if show_figures and not debug_kwargs:
+            for item in self.figures:
+                box_stack.append(item.bbox)
+                category_names_list.append(item.category_name.value)
+
         if show_tables and not debug_kwargs:
             for table in self.tables:
                 box_stack.append(table.bbox)
@@ -914,24 +979,34 @@ class Page(Image):
                         else:
                             category_names_list.append(word.token_tag.value if word.token_tag is not None else None)
 
+        if show_residual_layouts and not debug_kwargs:
+            for item in self.residual_layouts:
+                box_stack.append(item.bbox)
+                category_names_list.append(item.category_name.value)
+
         if self.image is not None:
+            scale_fx = scaled_width / self.width
+            scaled_height = int(self.height * scale_fx)
+            img = viz_handler.resize(self.image, scaled_width, scaled_height, "VIZ")
+
             if box_stack:
                 boxes = np.vstack(box_stack)
+                boxes = box_to_point4(boxes)
+                resizer = ResizeTransform(self.height, self.width, scaled_height, scaled_width, "VIZ")
+                boxes = resizer.apply_coords(boxes)
+                boxes = point4_to_box(boxes)
                 if show_words:
                     img = draw_boxes(
-                        self.image,
-                        boxes,
-                        category_names_list,
+                        np_image=img,
+                        boxes=boxes,
+                        category_names_list=category_names_list,
                         font_scale=1.0,
                         rectangle_thickness=4,
                     )
                 else:
-                    img = draw_boxes(self.image, boxes, category_names_list)
-                scale_fx, scale_fy = 1.3, 1.3
-                scaled_width, scaled_height = int(self.width * scale_fx), int(self.height * scale_fy)
-                img = viz_handler.resize(img, scaled_width, scaled_height, "VIZ")
-            else:
-                img = self.image
+                    img = draw_boxes(
+                        np_image=img, boxes=boxes, category_names_list=category_names_list, show_palette=False
+                    )
 
             if interactive:
                 interactive_imshow(img)

deepdoctection/eval/cocometric.py

@@ -71,8 +71,8 @@ https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocotools/cocoeva
 
 
 def _summarize(  # type: ignore
-    self, ap: int = 1, iouThr: float = 0.9, areaRng: str = "all", maxDets: int = 100
-) -> float:
+    self, ap: int = 1, iouThr: float = 0.9, areaRng: str = "all", maxDets: int = 100, per_category: bool = False
+) -> Union[float, list[float]]:
     # pylint: disable=C0103
     p = self.params
     iStr = " {:<18} {} @[ IoU={:<9} | area={:>6s} | maxDets={:>3d} ] = {:0.3f}"
@@ -86,6 +86,36 @@ def _summarize(  # type: ignore
 
     aind = [i for i, aRng in enumerate(p.areaRngLbl) if aRng == areaRng]
     mind = [i for i, mDet in enumerate(p.maxDets) if mDet == maxDets]
+    if per_category:
+        if ap == 1:
+            s = self.eval["precision"]
+            num_classes = s.shape[2]
+            results_per_class = []
+            for idx in range(num_classes):
+                if iouThr is not None:
+                    s = self.eval["precision"]
+                    t = np.where(iouThr == p.iouThrs)[0]
+                    s = s[t]
+                precision = s[:, :, idx, aind, mind]
+                precision = precision[precision > -1]
+                res = np.mean(precision) if precision.size else float("nan")
+                results_per_class.append(float(res))
+                print(f"Precision for class {idx+1}: @[ IoU={iouStr} | area={areaRng} | maxDets={maxDets} ] = {res}")
+        else:
+            s = self.eval["recall"]
+            num_classes = s.shape[1]
+            results_per_class = []
+            for idx in range(num_classes):
+                if iouThr is not None:
+                    s = self.eval["recall"]
+                    t = np.where(iouThr == p.iouThrs)[0]
+                    s = s[t]
+                recall = s[:, idx, aind, mind]
+                recall = recall[recall > -1]
+                res = np.mean(recall) if recall.size else float("nan")
+                results_per_class.append(float(res))
+                print(f"Recall for class {idx+1}: @[ IoU={iouStr} | area={areaRng} | maxDets={maxDets} ] = {res}")
+        return results_per_class
     if ap == 1:
         # dimension of precision: [TxRxKxAxM]
         s = self.eval["precision"]
@@ -124,6 +154,7 @@ class CocoMetric(MetricBase):
     mapper = image_to_coco
     _f1_score = None
     _f1_iou = None
+    _per_category = False
     _params: dict[str, Union[list[int], list[list[int]]]] = {}
 
     @classmethod
@@ -176,18 +207,28 @@ class CocoMetric(MetricBase):
 
         if cls._f1_score:
             summary_bbox = [
-                metric.summarize_f1(1, cls._f1_iou, maxDets=metric.params.maxDets[2]),
-                metric.summarize_f1(0, cls._f1_iou, maxDets=metric.params.maxDets[2]),
+                metric.summarize_f1(1, cls._f1_iou, maxDets=metric.params.maxDets[2], per_category=cls._per_category),
+                metric.summarize_f1(0, cls._f1_iou, maxDets=metric.params.maxDets[2], per_category=cls._per_category),
             ]
         else:
             metric.summarize()
             summary_bbox = metric.stats
 
         results = []
-        for params, value in zip(cls.get_summary_default_parameters(), summary_bbox):
+
+        default_parameters = cls.get_summary_default_parameters()
+        if cls._per_category:
+            default_parameters = default_parameters * len(summary_bbox[0])
+            summary_bbox = [item for pair in zip(*summary_bbox) for item in pair]
+        val = 0
+        for idx, (params, value) in enumerate(zip(default_parameters, summary_bbox)):
             params = copy(params)
             params["mode"] = "bbox"
             params["val"] = value
+            if cls._per_category:
+                if idx % 2 == 0:
+                    val += 1
+                params["category_id"] = val
             results.append(params)
 
         return results
@@ -201,15 +242,16 @@ class CocoMetric(MetricBase):
                  area range and maximum detections.
         """
         if cls._f1_score:
+            for el, idx in zip(_F1_DEFAULTS, [2, 2]):
+                if cls._params:
+                    if cls._params.get("maxDets") is not None:
+                        el["maxDets"] = cls._params["maxDets"][idx]
+                el["iouThr"] = cls._f1_iou
+            return _F1_DEFAULTS
+
+        for el, idx in zip(_COCOEVAL_DEFAULTS, _MAX_DET_INDEX):
             if cls._params:
                 if cls._params.get("maxDets") is not None:
-                    for el, idx in zip(_F1_DEFAULTS, [2, 2]):
-                        el["maxDets"] = cls._params["maxDets"][idx]
-                        el["iouThr"] = cls._f1_iou
-                    return _F1_DEFAULTS
-        if cls._params:
-            if cls._params.get("maxDets") is not None:
-                for el, idx in zip(_COCOEVAL_DEFAULTS, _MAX_DET_INDEX):
                     el["maxDets"] = cls._params["maxDets"][idx]
         return _COCOEVAL_DEFAULTS
 
@@ -220,13 +262,16 @@ class CocoMetric(MetricBase):
         area_range: Optional[list[list[int]]] = None,
         f1_score: bool = False,
         f1_iou: float = 0.9,
+        per_category: bool = False,
     ) -> None:
         """
         Setting params for different coco metric modes.
 
         :param max_detections: The maximum number of detections to consider
         :param area_range: The area range to classify objects as "all", "small", "medium" and "large"
-        :param f1_score: Will use f1 score setting with default iouThr 0.9
+        :param f1_score: Will use f1 score setting with default iouThr 0.9. To be more precise it does not calculate
+                         the f1 score but the precision and recall for a given iou threshold. Use the harmonic mean to
+                         get the ultimate f1 score.
         :param f1_iou: Use with f1_score True and reset the f1 iou threshold
         """
         if max_detections is not None:
@@ -238,6 +283,7 @@ class CocoMetric(MetricBase):
 
         cls._f1_score = f1_score
         cls._f1_iou = f1_iou
+        cls._per_category = per_category
 
     @classmethod
     def get_requirements(cls) -> list[Requirement]:

deepdoctection/mapper/match.py

@@ -193,5 +193,7 @@ def match_anns_by_distance(
     child_anns = dp.get_annotation(annotation_ids=child_ann_ids, category_names=child_ann_category_names)
     child_centers = [block.get_bounding_box(dp.image_id).center for block in child_anns]
     parent_centers = [block.get_bounding_box(dp.image_id).center for block in parent_anns]
-    child_indices = distance.cdist(parent_centers, child_centers).argmin(axis=1)
-    return [(parent_anns[i], child_anns[j]) for i, j in enumerate(child_indices)]
+    if child_centers and parent_centers:
+        child_indices = distance.cdist(parent_centers, child_centers).argmin(axis=1)
+        return [(parent_anns[i], child_anns[j]) for i, j in enumerate(child_indices)]
+    return []

deepdoctection/utils/utils.py

@@ -155,3 +155,42 @@ def is_file_extension(file_name: PathLikeOrStr, extension: Union[str, Sequence[s
     if isinstance(extension, str):
         return os.path.splitext(file_name)[-1].lower() == extension
     return os.path.splitext(file_name)[-1].lower() in extension
+
+
+def partition_list(base_list: list[str], stop_value: str) -> list[list[str]]:
+    """
+    Partitions a list of strings into sublists, where each sublist starts with the first occurrence of the stop value.
+    Consecutive stop values are grouped together in the same sublist.
+
+    :param base_list: The list of strings to be partitioned.
+    :param stop_value: The string value that indicates the start of a new partition.
+    :return: A list of lists, where each sublist is a partition of the original list.
+
+    ** Example:**
+
+        strings = ['a', 'a', 'c', 'c', 'b', 'd', 'c', 'c', 'a', 'b', 'a', 'b', 'a', 'a']
+        stop_string = 'a'
+        partition_list(strings, stop_string)
+
+       # Output [['a', 'a', 'c', 'c', 'b', 'd', 'c', 'c'], ['a', 'b'], ['a', 'b'], ['a', 'a']]
+    """
+
+    partitions = []
+    current_partition: list[str] = []
+    stop_found = False
+
+    for s in base_list:
+        if s == stop_value:
+            if not stop_found and current_partition:
+                partitions.append(current_partition)
+                current_partition = []
+            current_partition.append(s)
+            stop_found = True
+        else:
+            current_partition.append(s)
+            stop_found = False
+
+    if current_partition:
+        partitions.append(current_partition)
+
+    return partitions

deepdoctection/utils/viz.py

@@ -205,6 +205,7 @@ def draw_boxes(
     font_scale: float = 1.0,
     rectangle_thickness: int = 4,
     box_color_by_category: bool = True,
+    show_palette: bool = True,
 ) -> PixelValues:
     """
     Dray bounding boxes with category names into image.
@@ -216,6 +217,7 @@ def draw_boxes(
     :param font_scale: Font scale of text box
     :param rectangle_thickness: Thickness of bounding box
     :param box_color_by_category:
+    :param show_palette: Whether to show a color palette of the categories
     :return: A new image np.ndarray
     """
     if color is not None:
@@ -261,19 +263,20 @@ def draw_boxes(
         )
 
     # draw a (very ugly) color palette
-    y_0 = np_image.shape[0]
-    for category, col in category_to_color.items():
-        if category is not None:
-            np_image = viz_handler.draw_text(
-                np_image,
-                (np_image.shape[1], y_0),
-                category,
-                color=col,
-                font_scale=font_scale * 3,
-                rectangle_thickness=rectangle_thickness,
-            )
-            _, text_h = viz_handler.get_text_size(category, font_scale * 2)
-            y_0 = y_0 - int(10 * text_h)
+    if show_palette:
+        y_0 = np_image.shape[0]
+        for category, col in category_to_color.items():
+            if category is not None:
+                np_image = viz_handler.draw_text(
+                    np_image,
+                    (np_image.shape[1], y_0),
+                    category,
+                    color=col,
+                    font_scale=font_scale,
+                    rectangle_thickness=rectangle_thickness,
+                )
+                _, text_h = viz_handler.get_text_size(category, font_scale * 2)
+                y_0 = y_0 - int(1 * text_h)
 
     return np_image
 

{deepdoctection-0.35.dist-info → deepdoctection-0.36.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: deepdoctection
-Version: 0.35
+Version: 0.36
 Summary: Repository for Document AI
 Home-page: https://github.com/deepdoctection/deepdoctection
 Author: Dr. Janis Meyer
@@ -16,117 +16,117 @@ Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Requires-Python: >=3.9
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: catalogue ==2.0.10
-Requires-Dist: huggingface-hub <0.26,>=0.12.0
-Requires-Dist: importlib-metadata >=5.0.0
-Requires-Dist: jsonlines ==3.1.0
-Requires-Dist: lazy-imports ==0.3.1
-Requires-Dist: mock ==4.0.3
-Requires-Dist: networkx >=2.7.1
-Requires-Dist: numpy <2.0,>=1.21
-Requires-Dist: packaging >=20.0
-Requires-Dist: Pillow >=10.0.0
-Requires-Dist: pypdf >=3.16.0
-Requires-Dist: pypdfium2 >=4.30.0
-Requires-Dist: pyyaml >=6.0.1
-Requires-Dist: pyzmq >=16
-Requires-Dist: scipy >=1.13.1
-Requires-Dist: termcolor >=1.1
-Requires-Dist: tabulate >=0.7.7
-Requires-Dist: tqdm ==4.64.0
-Provides-Extra: dev
-Requires-Dist: python-dotenv ==1.0.0 ; extra == 'dev'
-Requires-Dist: click ; extra == 'dev'
-Requires-Dist: black ==23.7.0 ; extra == 'dev'
-Requires-Dist: isort ==5.13.2 ; extra == 'dev'
-Requires-Dist: pylint ==2.17.4 ; extra == 'dev'
-Requires-Dist: mypy ==1.4.1 ; extra == 'dev'
-Requires-Dist: wandb ; extra == 'dev'
-Requires-Dist: types-PyYAML >=6.0.12.12 ; extra == 'dev'
-Requires-Dist: types-termcolor >=1.1.3 ; extra == 'dev'
-Requires-Dist: types-tabulate >=0.9.0.3 ; extra == 'dev'
-Requires-Dist: types-tqdm >=4.66.0.5 ; extra == 'dev'
-Requires-Dist: lxml-stubs >=0.5.1 ; extra == 'dev'
-Requires-Dist: types-Pillow >=10.2.0.20240406 ; extra == 'dev'
-Requires-Dist: types-urllib3 >=1.26.25.14 ; extra == 'dev'
-Provides-Extra: docs
-Requires-Dist: tensorpack ==0.11 ; extra == 'docs'
-Requires-Dist: boto3 ==1.34.102 ; extra == 'docs'
-Requires-Dist: transformers >=4.36.0 ; extra == 'docs'
-Requires-Dist: accelerate >=0.29.1 ; extra == 'docs'
-Requires-Dist: pdfplumber >=0.11.0 ; extra == 'docs'
-Requires-Dist: lxml >=4.9.1 ; extra == 'docs'
-Requires-Dist: lxml-stubs >=0.5.1 ; extra == 'docs'
-Requires-Dist: jdeskew >=0.2.2 ; extra == 'docs'
-Requires-Dist: jinja2 ==3.0.3 ; extra == 'docs'
-Requires-Dist: mkdocs-material ; extra == 'docs'
-Requires-Dist: mkdocstrings-python ; extra == 'docs'
-Requires-Dist: griffe ==0.25.0 ; extra == 'docs'
+Requires-Dist: catalogue==2.0.10
+Requires-Dist: huggingface_hub<0.26,>=0.12.0
+Requires-Dist: importlib-metadata>=5.0.0
+Requires-Dist: jsonlines==3.1.0
+Requires-Dist: lazy-imports==0.3.1
+Requires-Dist: mock==4.0.3
+Requires-Dist: networkx>=2.7.1
+Requires-Dist: numpy<2.0,>=1.21
+Requires-Dist: packaging>=20.0
+Requires-Dist: Pillow>=10.0.0
+Requires-Dist: pypdf>=3.16.0
+Requires-Dist: pypdfium2>=4.30.0
+Requires-Dist: pyyaml>=6.0.1
+Requires-Dist: pyzmq>=16
+Requires-Dist: scipy>=1.13.1
+Requires-Dist: termcolor>=1.1
+Requires-Dist: tabulate>=0.7.7
+Requires-Dist: tqdm==4.64.0
+Provides-Extra: tf
+Requires-Dist: catalogue==2.0.10; extra == "tf"
+Requires-Dist: huggingface_hub<0.26,>=0.12.0; extra == "tf"
+Requires-Dist: importlib-metadata>=5.0.0; extra == "tf"
+Requires-Dist: jsonlines==3.1.0; extra == "tf"
+Requires-Dist: lazy-imports==0.3.1; extra == "tf"
+Requires-Dist: mock==4.0.3; extra == "tf"
+Requires-Dist: networkx>=2.7.1; extra == "tf"
+Requires-Dist: numpy<2.0,>=1.21; extra == "tf"
+Requires-Dist: packaging>=20.0; extra == "tf"
+Requires-Dist: Pillow>=10.0.0; extra == "tf"
+Requires-Dist: pypdf>=3.16.0; extra == "tf"
+Requires-Dist: pypdfium2>=4.30.0; extra == "tf"
+Requires-Dist: pyyaml>=6.0.1; extra == "tf"
+Requires-Dist: pyzmq>=16; extra == "tf"
+Requires-Dist: scipy>=1.13.1; extra == "tf"
+Requires-Dist: termcolor>=1.1; extra == "tf"
+Requires-Dist: tabulate>=0.7.7; extra == "tf"
+Requires-Dist: tqdm==4.64.0; extra == "tf"
+Requires-Dist: tensorpack==0.11; extra == "tf"
+Requires-Dist: protobuf==3.20.1; extra == "tf"
+Requires-Dist: tensorflow-addons>=0.17.1; extra == "tf"
+Requires-Dist: tf2onnx>=1.9.2; extra == "tf"
+Requires-Dist: python-doctr==0.8.1; extra == "tf"
+Requires-Dist: pycocotools>=2.0.2; extra == "tf"
+Requires-Dist: boto3==1.34.102; extra == "tf"
+Requires-Dist: pdfplumber>=0.11.0; extra == "tf"
+Requires-Dist: fasttext==0.9.2; extra == "tf"
+Requires-Dist: jdeskew>=0.2.2; extra == "tf"
+Requires-Dist: apted==1.0.3; extra == "tf"
+Requires-Dist: distance==0.1.3; extra == "tf"
+Requires-Dist: lxml>=4.9.1; extra == "tf"
 Provides-Extra: pt
-Requires-Dist: catalogue ==2.0.10 ; extra == 'pt'
-Requires-Dist: huggingface-hub <0.26,>=0.12.0 ; extra == 'pt'
-Requires-Dist: importlib-metadata >=5.0.0 ; extra == 'pt'
-Requires-Dist: jsonlines ==3.1.0 ; extra == 'pt'
-Requires-Dist: lazy-imports ==0.3.1 ; extra == 'pt'
-Requires-Dist: mock ==4.0.3 ; extra == 'pt'
-Requires-Dist: networkx >=2.7.1 ; extra == 'pt'
-Requires-Dist: numpy <2.0,>=1.21 ; extra == 'pt'
-Requires-Dist: packaging >=20.0 ; extra == 'pt'
-Requires-Dist: Pillow >=10.0.0 ; extra == 'pt'
-Requires-Dist: pypdf >=3.16.0 ; extra == 'pt'
-Requires-Dist: pypdfium2 >=4.30.0 ; extra == 'pt'
-Requires-Dist: pyyaml >=6.0.1 ; extra == 'pt'
-Requires-Dist: pyzmq >=16 ; extra == 'pt'
-Requires-Dist: scipy >=1.13.1 ; extra == 'pt'
-Requires-Dist: termcolor >=1.1 ; extra == 'pt'
-Requires-Dist: tabulate >=0.7.7 ; extra == 'pt'
-Requires-Dist: tqdm ==4.64.0 ; extra == 'pt'
-Requires-Dist: timm >=0.9.16 ; extra == 'pt'
-Requires-Dist: transformers >=4.36.0 ; extra == 'pt'
-Requires-Dist: accelerate >=0.29.1 ; extra == 'pt'
-Requires-Dist: python-doctr ==0.8.1 ; extra == 'pt'
-Requires-Dist: boto3 ==1.34.102 ; extra == 'pt'
-Requires-Dist: pdfplumber >=0.11.0 ; extra == 'pt'
-Requires-Dist: fasttext ==0.9.2 ; extra == 'pt'
-Requires-Dist: jdeskew >=0.2.2 ; extra == 'pt'
-Requires-Dist: apted ==1.0.3 ; extra == 'pt'
-Requires-Dist: distance ==0.1.3 ; extra == 'pt'
-Requires-Dist: lxml >=4.9.1 ; extra == 'pt'
+Requires-Dist: catalogue==2.0.10; extra == "pt"
+Requires-Dist: huggingface_hub<0.26,>=0.12.0; extra == "pt"
+Requires-Dist: importlib-metadata>=5.0.0; extra == "pt"
+Requires-Dist: jsonlines==3.1.0; extra == "pt"
+Requires-Dist: lazy-imports==0.3.1; extra == "pt"
+Requires-Dist: mock==4.0.3; extra == "pt"
+Requires-Dist: networkx>=2.7.1; extra == "pt"
+Requires-Dist: numpy<2.0,>=1.21; extra == "pt"
+Requires-Dist: packaging>=20.0; extra == "pt"
+Requires-Dist: Pillow>=10.0.0; extra == "pt"
+Requires-Dist: pypdf>=3.16.0; extra == "pt"
+Requires-Dist: pypdfium2>=4.30.0; extra == "pt"
+Requires-Dist: pyyaml>=6.0.1; extra == "pt"
+Requires-Dist: pyzmq>=16; extra == "pt"
+Requires-Dist: scipy>=1.13.1; extra == "pt"
+Requires-Dist: termcolor>=1.1; extra == "pt"
+Requires-Dist: tabulate>=0.7.7; extra == "pt"
+Requires-Dist: tqdm==4.64.0; extra == "pt"
+Requires-Dist: timm>=0.9.16; extra == "pt"
+Requires-Dist: transformers>=4.36.0; extra == "pt"
+Requires-Dist: accelerate>=0.29.1; extra == "pt"
+Requires-Dist: python-doctr==0.8.1; extra == "pt"
+Requires-Dist: boto3==1.34.102; extra == "pt"
+Requires-Dist: pdfplumber>=0.11.0; extra == "pt"
+Requires-Dist: fasttext==0.9.2; extra == "pt"
+Requires-Dist: jdeskew>=0.2.2; extra == "pt"
+Requires-Dist: apted==1.0.3; extra == "pt"
+Requires-Dist: distance==0.1.3; extra == "pt"
+Requires-Dist: lxml>=4.9.1; extra == "pt"
+Provides-Extra: docs
+Requires-Dist: tensorpack==0.11; extra == "docs"
+Requires-Dist: boto3==1.34.102; extra == "docs"
+Requires-Dist: transformers>=4.36.0; extra == "docs"
+Requires-Dist: accelerate>=0.29.1; extra == "docs"
+Requires-Dist: pdfplumber>=0.11.0; extra == "docs"
+Requires-Dist: lxml>=4.9.1; extra == "docs"
+Requires-Dist: lxml-stubs>=0.5.1; extra == "docs"
+Requires-Dist: jdeskew>=0.2.2; extra == "docs"
+Requires-Dist: jinja2==3.0.3; extra == "docs"
+Requires-Dist: mkdocs-material; extra == "docs"
+Requires-Dist: mkdocstrings-python; extra == "docs"
+Requires-Dist: griffe==0.25.0; extra == "docs"
+Provides-Extra: dev
+Requires-Dist: python-dotenv==1.0.0; extra == "dev"
+Requires-Dist: click; extra == "dev"
+Requires-Dist: black==23.7.0; extra == "dev"
+Requires-Dist: isort==5.13.2; extra == "dev"
+Requires-Dist: pylint==2.17.4; extra == "dev"
+Requires-Dist: mypy==1.4.1; extra == "dev"
+Requires-Dist: wandb; extra == "dev"
+Requires-Dist: types-PyYAML>=6.0.12.12; extra == "dev"
+Requires-Dist: types-termcolor>=1.1.3; extra == "dev"
+Requires-Dist: types-tabulate>=0.9.0.3; extra == "dev"
+Requires-Dist: types-tqdm>=4.66.0.5; extra == "dev"
+Requires-Dist: lxml-stubs>=0.5.1; extra == "dev"
+Requires-Dist: types-Pillow>=10.2.0.20240406; extra == "dev"
+Requires-Dist: types-urllib3>=1.26.25.14; extra == "dev"
 Provides-Extra: test
-Requires-Dist: pytest ==8.0.2 ; extra == 'test'
-Requires-Dist: pytest-cov ; extra == 'test'
-Provides-Extra: tf
-Requires-Dist: catalogue ==2.0.10 ; extra == 'tf'
-Requires-Dist: huggingface-hub <0.26,>=0.12.0 ; extra == 'tf'
-Requires-Dist: importlib-metadata >=5.0.0 ; extra == 'tf'
-Requires-Dist: jsonlines ==3.1.0 ; extra == 'tf'
-Requires-Dist: lazy-imports ==0.3.1 ; extra == 'tf'
-Requires-Dist: mock ==4.0.3 ; extra == 'tf'
-Requires-Dist: networkx >=2.7.1 ; extra == 'tf'
-Requires-Dist: numpy <2.0,>=1.21 ; extra == 'tf'
-Requires-Dist: packaging >=20.0 ; extra == 'tf'
-Requires-Dist: Pillow >=10.0.0 ; extra == 'tf'
-Requires-Dist: pypdf >=3.16.0 ; extra == 'tf'
-Requires-Dist: pypdfium2 >=4.30.0 ; extra == 'tf'
-Requires-Dist: pyyaml >=6.0.1 ; extra == 'tf'
-Requires-Dist: pyzmq >=16 ; extra == 'tf'
-Requires-Dist: scipy >=1.13.1 ; extra == 'tf'
-Requires-Dist: termcolor >=1.1 ; extra == 'tf'
-Requires-Dist: tabulate >=0.7.7 ; extra == 'tf'
-Requires-Dist: tqdm ==4.64.0 ; extra == 'tf'
-Requires-Dist: tensorpack ==0.11 ; extra == 'tf'
-Requires-Dist: protobuf ==3.20.1 ; extra == 'tf'
-Requires-Dist: tensorflow-addons >=0.17.1 ; extra == 'tf'
-Requires-Dist: tf2onnx >=1.9.2 ; extra == 'tf'
-Requires-Dist: python-doctr ==0.8.1 ; extra == 'tf'
-Requires-Dist: pycocotools >=2.0.2 ; extra == 'tf'
-Requires-Dist: boto3 ==1.34.102 ; extra == 'tf'
-Requires-Dist: pdfplumber >=0.11.0 ; extra == 'tf'
-Requires-Dist: fasttext ==0.9.2 ; extra == 'tf'
-Requires-Dist: jdeskew >=0.2.2 ; extra == 'tf'
-Requires-Dist: apted ==1.0.3 ; extra == 'tf'
-Requires-Dist: distance ==0.1.3 ; extra == 'tf'
-Requires-Dist: lxml >=4.9.1 ; extra == 'tf'
+Requires-Dist: pytest==8.0.2; extra == "test"
+Requires-Dist: pytest-cov; extra == "test"
 
 
 <p align="center">
@@ -176,12 +176,16 @@ pipelines. Its core function does not depend on any specific deep learning libra
    [**Torchscript**](https://pytorch.org/docs/stable/jit.html) (CPU) as well and [**Detectron2**](https://github.com/facebookresearch/detectron2/tree/main/detectron2) is not required 
    anymore for basic inference.
  - More angle predictors for determining the rotation of a document based on [**Tesseract**](https://github.com/tesseract-ocr/tesseract) and [**DocTr**](https://github.com/mindee/doctr)
-   (not contained in the built-in Analyzer).
  - Token classification with [**LiLT**](https://github.com/jpWang/LiLT) via 
    [**transformers**](https://github.com/huggingface/transformers). 
    We have added a model wrapper for token classification with LiLT and added a some LiLT models to the model catalog 
    that seem to look promising, especially if you want to train a model on non-english data. The training script for 
-   LayoutLM can be used for LiLT as well and we will be providing a notebook on how to train a model on a custom dataset soon.
+   LayoutLM can be used for LiLT as well.
+ - [**new**] There are two notebooks available that show, how to write a 
+   [custom predictor](https://github.com/deepdoctection/notebooks/blob/main/Doclaynet_Analyzer_Config.ipynb) based on 
+   a third party library that has not been supported yet and how to use 
+   [advanced configuration](https://github.com/deepdoctection/notebooks/blob/main/Doclaynet_Analyzer_Config.ipynb) to 
+   get links between layout segments e.g. captions and tables or figures.
 
 **deep**doctection provides on top of that methods for pre-processing inputs to models like cropping or resizing and to 
 post-process results, like validating duplicate outputs, relating words to detected layout segments or ordering words 

{deepdoctection-0.35.dist-info → deepdoctection-0.36.dist-info}/RECORD

@@ -1,11 +1,11 @@
-deepdoctection/__init__.py,sha256=RZpawNRTJPKNPFuONawVOsYWdr-rI8PPNXZhlPtOKtc,12580
+deepdoctection/__init__.py,sha256=fNUbaFAlK1JUXgPCmTu2UOLUMqW4HIgkaW4uOUYjYYg,12571
 deepdoctection/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 deepdoctection/analyzer/__init__.py,sha256=icClxrd20XutD6LxLgEPIWceSs4j_QfI3szCE-9BL2w,729
-deepdoctection/analyzer/_config.py,sha256=0cWtaI2e3jHNhufHZAqMje0YTTDAogKAHVl4VpYojAo,4874
+deepdoctection/analyzer/_config.py,sha256=NZl_REM8Ge2xfxvHN-mZR5KURcHfZii3xfMlKQwckbA,4864
 deepdoctection/analyzer/dd.py,sha256=DUOhOtwipHw5nabYqn3WGR9aZcgP0ma_bi_tjf9xscw,5973
-deepdoctection/analyzer/factory.py,sha256=T9jxtVLNFhocbsfWIGLPfFrEv21zQJzM6VdFt0yxMyg,23849
+deepdoctection/analyzer/factory.py,sha256=xmo5F9X7I6lp0ZWJv8QavpMyG8UWYLvMi4qogsZV1_s,31507
 deepdoctection/configs/__init__.py,sha256=TX_P6tqDOF1LK1mi9ruAl7x0mtv1Asm8cYWCz3Pe2dk,646
-deepdoctection/configs/conf_dd_one.yaml,sha256=orP-oeqtWbz5S9FJZJKxy1UqMwOYjL9g0DOX-wbamqU,2239
+deepdoctection/configs/conf_dd_one.yaml,sha256=td7XsyVhdXkhh5Pie7sT_WNjGTaxBOWgpxhkobHd1H0,2325
 deepdoctection/configs/conf_tesseract.yaml,sha256=oF6szDyoi15FHvq7yFUNIEjfA_jNLhGxoowiRsz_zY4,35
 deepdoctection/dataflow/__init__.py,sha256=CWRHMpmJaPk4xY_oIIFubCt-z11SguWrMWxHZ7rdrvY,845
 deepdoctection/dataflow/base.py,sha256=z4DCComSj5wStEPjtk0093cNNGfUMiDqx8dqz36nS_o,6221
@@ -20,7 +20,7 @@ deepdoctection/datapoint/annotation.py,sha256=FEgz4COxVDfjic0gG7kS6iHnWLBIgFnquQ
 deepdoctection/datapoint/box.py,sha256=tkFuVM6xfx2jL7W4UED4qHXV572LSRdIsVJbrEiyIxI,23524
 deepdoctection/datapoint/convert.py,sha256=Gw2IjNiEotPu1yuMZqrIYB0mCAwafKt-VgMnrHj6S7U,6808
 deepdoctection/datapoint/image.py,sha256=EvZlVwJjMAcL1z8RNPBvZ8fwdJvkGuGpcFxCP1y26Go,33045
-deepdoctection/datapoint/view.py,sha256=7qSX4DQw9OPQQSKfSjV8e5i6jLyu6hOMceSKJAob2N8,42154
+deepdoctection/datapoint/view.py,sha256=1rVMuqucCrI5zlwyXMADJQBV38V_zSNFqFyBi3cMA1E,44914
 deepdoctection/datasets/__init__.py,sha256=-A3aR90aDsHPmVM35JavfnQ2itYSCn3ujl4krRni1QU,1076
 deepdoctection/datasets/adapter.py,sha256=Ly_vbOAgVI73V41FUccnSX1ECTOyesW_qsuvQuvOZbw,7796
 deepdoctection/datasets/base.py,sha256=DT4i-d74sIEiUNC6UspIHNJuHSK0t1dBv7qwadg4rLw,22341
@@ -44,7 +44,7 @@ deepdoctection/datasets/instances/xsl/pascal_voc.xsl,sha256=DlzFV2P8NtQKXVe96i-m
 deepdoctection/eval/__init__.py,sha256=rbns4tSEQ30QLj8h0mm3A0dCaKuN9LDxxpVypKKSXSE,932
 deepdoctection/eval/accmetric.py,sha256=4bND-xz9AZu9ACYRkEzn9V6Jn8MEiqnF7kxSp4k_baE,19655
 deepdoctection/eval/base.py,sha256=gCvhTdwEaCKplYTWPMjGvtB_0Vbq2KBJWFHq8mMlLPA,4814
-deepdoctection/eval/cocometric.py,sha256=Co7XaLQzp7qxw8UQaG2D68PzY6eA9aRNueeo_zaMJLM,8777
+deepdoctection/eval/cocometric.py,sha256=4cpNmF3xZjInCOWOoVU_7itQxLI-zr0O6suNjPU2xWc,11020
 deepdoctection/eval/eval.py,sha256=B9PUZBjj6KzXHLOxUVn3QHiOcBQogfJmp9mjopbMo9k,19721
 deepdoctection/eval/registry.py,sha256=v4mp-s67vBVRu1nQzuGlYPViQnMSeIXEcF_WmvfUCoU,1051
 deepdoctection/eval/tedsmetric.py,sha256=rKw-734Y9CpBtIfkBSPQF2vAZxnIdWrI9Zc723P7RxI,9529
@@ -94,7 +94,7 @@ deepdoctection/mapper/d2struct.py,sha256=Dx-YnycsIQH4a5-9Gn_yMhiQ-gOFgMueNeH3rhX
 deepdoctection/mapper/hfstruct.py,sha256=2PjGKsYturVJBimLT1CahYh09KSRAFEHz_QNtC162kQ,5551
 deepdoctection/mapper/laylmstruct.py,sha256=abMZkYU2W0e_VcCm_c0ZXNFuv-lfMFWcTedcZS5EYvE,42935
 deepdoctection/mapper/maputils.py,sha256=eI6ZcDg9W5uB6xQNBZpMIdEd86HlCxTtkJuyROdTqiw,8146
-deepdoctection/mapper/match.py,sha256=E7Qna6zLDIxlI7puOL9BjjZKuRry-zONs8TLWmyEMIQ,9580
+deepdoctection/mapper/match.py,sha256=pCWZpz2R8JahiKXCw7dxKRTLiPgJXeVDgkddDPLy_c0,9643
 deepdoctection/mapper/misc.py,sha256=rCqHOcsCfVPXs36AWK0rZ2kk0CUM3yXV370_zyIGBJ4,6518
 deepdoctection/mapper/pascalstruct.py,sha256=TzVU1p0oiw0nOuxTFFbEB9vXJxH1v6VUvTJ7MD0manU,3828
 deepdoctection/mapper/prodigystruct.py,sha256=Re4Sd_zAp6qOvbXZLmMJeG0IGEfMQxebuyDeZgMcTa8,6827
@@ -139,10 +139,10 @@ deepdoctection/utils/settings.py,sha256=k6OyuWbj-IPeaO9zT9RZ-5Yad1wNhWGYqGLZdtgX
 deepdoctection/utils/tqdm.py,sha256=cBUtR0L1x0KMeYrLP2rrzyzCamCjpQAKroHXLv81_pk,1820
 deepdoctection/utils/transform.py,sha256=3kCgsEeRkG1efCdkfvj7tUFMs-e2jbjbflq826F2GPU,8502
 deepdoctection/utils/types.py,sha256=_3dmPdCIZNLbgU5QP5k_c5phDf18xLe1kYL6t2nM45s,2953
-deepdoctection/utils/utils.py,sha256=ANzyIX6AY1yc-4gcn6yxksV84sPrJDaUurUNVatAFu8,5168
-deepdoctection/utils/viz.py,sha256=Xm6pKlhM29UWBBGZHlWFl9XYFDAqaYDdwHXwe26Hvqo,25728
-deepdoctection-0.35.dist-info/LICENSE,sha256=GQ0rUvuGdrMNEI3iHK5UQx6dIMU1QwAuyXsxUHn5MEQ,11351
-deepdoctection-0.35.dist-info/METADATA,sha256=B6pPQjRYWcqd1p-3ul3PhflYOcKq2ZpP5D-i8kr7qgk,19403
-deepdoctection-0.35.dist-info/WHEEL,sha256=P9jw-gEje8ByB7_hXoICnHtVCrEwMQh-630tKvQWehc,91
-deepdoctection-0.35.dist-info/top_level.txt,sha256=hs2DdoOL9h4mnHhmO82BT4pz4QATIoOZ20PZmlnxFI8,15
-deepdoctection-0.35.dist-info/RECORD,,
+deepdoctection/utils/utils.py,sha256=csVs_VvCq4QBETPoE2JdTTL4MFYnD4xh-Js5vRb612g,6492
+deepdoctection/utils/viz.py,sha256=Mok1d0V7NwlhAvO1S1Iq5YitKpVmOfH_XHTSlRelCB0,25902
+deepdoctection-0.36.dist-info/LICENSE,sha256=GQ0rUvuGdrMNEI3iHK5UQx6dIMU1QwAuyXsxUHn5MEQ,11351
+deepdoctection-0.36.dist-info/METADATA,sha256=E-zXgx0bTdSqbd88D_abscR_poEJaKJGIwlv2RFbQs8,19543
+deepdoctection-0.36.dist-info/WHEEL,sha256=PZUExdf71Ui_so67QXpySuHtCi3-J3wvF4ORK6k_S8U,91
+deepdoctection-0.36.dist-info/top_level.txt,sha256=hs2DdoOL9h4mnHhmO82BT4pz4QATIoOZ20PZmlnxFI8,15
+deepdoctection-0.36.dist-info/RECORD,,

{deepdoctection-0.35.dist-info → deepdoctection-0.36.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (75.3.0)
+Generator: setuptools (75.6.0)
 Root-Is-Purelib: true
 Tag: py3-none-any