paddlex 3.0.3__py3-none-any.whl → 3.1.0__py3-none-any.whl

paddlex/inference/serving/schemas/pp_chatocrv3_doc.py

@@ -57,6 +57,7 @@ class AnalyzeImagesRequest(ocr.BaseInferRequest):
     sealDetBoxThresh: Optional[float] = None
     sealDetUnclipRatio: Optional[float] = None
     sealRecScoreThresh: Optional[float] = None
+    visualize: Optional[bool] = None
 
 
 class LayoutParsingResult(BaseModel):
@@ -78,8 +79,8 @@ BUILD_VECTOR_STORE_ENDPOINT: Final[str] = "/chatocr-vector"
 
 class BuildVectorStoreRequest(BaseModel):
     visualInfo: List[dict]
-    minCharacters: Optional[int] = None
-    blockSize: Optional[int] = None
+    minCharacters: int = 3500
+    blockSize: int = 300
     retrieverConfig: Optional[dict] = None
 
 
@@ -93,9 +94,9 @@ CHAT_ENDPOINT: Final[str] = "/chatocr-chat"
 class ChatRequest(BaseModel):
     keyList: List[str]
     visualInfo: List[dict]
-    useVectorRetrieval: Optional[bool] = None
+    useVectorRetrieval: bool = True
     vectorInfo: Optional[dict] = None
-    minCharacters: Optional[int] = None
+    minCharacters: int = 3500
     textTaskDescription: Optional[str] = None
     textOutputFormat: Optional[str] = None
     # Is the "Str" in the name unnecessary? Keep the names consistent with the

paddlex/inference/serving/schemas/pp_chatocrv4_doc.py

@@ -61,6 +61,7 @@ class AnalyzeImagesRequest(ocr.BaseInferRequest):
     sealDetBoxThresh: Optional[float] = None
     sealDetUnclipRatio: Optional[float] = None
     sealRecScoreThresh: Optional[float] = None
+    visualize: Optional[bool] = None
 
 
 class LayoutParsingResult(BaseModel):
@@ -80,8 +81,8 @@ BUILD_VECTOR_STORE_ENDPOINT: Final[str] = "/chatocr-vector"
 
 class BuildVectorStoreRequest(BaseModel):
     visualInfo: List[dict]
-    minCharacters: Optional[int] = None
-    blockSize: Optional[int] = None
+    minCharacters: int = 3500
+    blockSize: int = 300
     retrieverConfig: Optional[dict] = None
 
 
@@ -108,9 +109,9 @@ CHAT_ENDPOINT: Final[str] = "/chatocr-chat"
 class ChatRequest(BaseModel):
     keyList: List[str]
     visualInfo: List[dict]
-    useVectorRetrieval: Optional[bool] = None
+    useVectorRetrieval: bool = True
     vectorInfo: Optional[dict] = None
-    minCharacters: Optional[int] = None
+    minCharacters: int = 3500
     textTaskDescription: Optional[str] = None
     textOutputFormat: Optional[str] = None
     textRulesStr: Optional[str] = None
@@ -122,7 +123,7 @@ class ChatRequest(BaseModel):
     tableFewShotDemoTextContent: Optional[str] = None
     tableFewShotDemoKeyValueList: Optional[str] = None
     mllmPredictInfo: Optional[dict] = None
-    mllmIntegrationStrategy: Optional[str] = None
+    mllmIntegrationStrategy: str = "integration"
     chatBotConfig: Optional[dict] = None
     retrieverConfig: Optional[dict] = None
 

paddlex/inference/serving/schemas/pp_doctranslation.py

@@ -0,0 +1,115 @@
+# Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from typing import Dict, Final, List, Optional, Tuple, Union
+
+from pydantic import BaseModel
+
+from ..infra.models import DataInfo, PrimaryOperations
+from .shared import ocr
+
+__all__ = [
+    "ANALYZE_IMAGES_ENDPOINT",
+    "AnalyzeImagesRequest",
+    "LayoutParsingResult",
+    "AnalyzeImagesResult",
+    "TRANSLATE_ENDPOINT",
+    "TranslateRequest",
+    "TranslationResult",
+    "TranslateResult",
+    "PRIMARY_OPERATIONS",
+]
+
+ANALYZE_IMAGES_ENDPOINT: Final[str] = "/doctrans-visual"
+
+
+class AnalyzeImagesRequest(ocr.BaseInferRequest):
+    useDocOrientationClassify: Optional[bool] = False
+    useDocUnwarping: Optional[bool] = False
+    useTextlineOrientation: Optional[bool] = None
+    useSealRecognition: Optional[bool] = None
+    useTableRecognition: Optional[bool] = None
+    useFormulaRecognition: Optional[bool] = None
+    useChartRecognition: Optional[bool] = False
+    useRegionDetection: Optional[bool] = None
+    layoutThreshold: Optional[Union[float, dict]] = None
+    layoutNms: Optional[bool] = None
+    layoutUnclipRatio: Optional[Union[float, Tuple[float, float], dict]] = None
+    layoutMergeBboxesMode: Optional[Union[str, dict]] = None
+    textDetLimitSideLen: Optional[int] = None
+    textDetLimitType: Optional[str] = None
+    textDetThresh: Optional[float] = None
+    textDetBoxThresh: Optional[float] = None
+    textDetUnclipRatio: Optional[float] = None
+    textRecScoreThresh: Optional[float] = None
+    sealDetLimitSideLen: Optional[int] = None
+    sealDetLimitType: Optional[str] = None
+    sealDetThresh: Optional[float] = None
+    sealDetBoxThresh: Optional[float] = None
+    sealDetUnclipRatio: Optional[float] = None
+    sealRecScoreThresh: Optional[float] = None
+    useWiredTableCellsTransToHtml: bool = False
+    useWirelessTableCellsTransToHtml: bool = False
+    useTableOrientationClassify: bool = True
+    useOcrResultsWithTableCells: bool = True
+    useE2eWiredTableRecModel: bool = False
+    useE2eWirelessTableRecModel: bool = True
+    visualize: Optional[bool] = None
+
+
+class LayoutParsingResult(BaseModel):
+    prunedResult: dict
+    markdown: ocr.MarkdownData
+    outputImages: Optional[Dict[str, str]] = None
+    inputImage: Optional[str] = None
+
+
+class AnalyzeImagesResult(BaseModel):
+    layoutParsingResults: List[LayoutParsingResult]
+    dataInfo: DataInfo
+
+
+TRANSLATE_ENDPOINT: Final[str] = "/doctrans-translate"
+
+
+class TranslateRequest(BaseModel):
+    markdownList: List[ocr.MarkdownData]
+    targetLanguage: str = "zh"
+    chunkSize: int = 5000
+    taskDescription: Optional[str] = None
+    outputFormat: Optional[str] = None
+    rulesStr: Optional[str] = None
+    fewShotDemoTextContent: Optional[str] = None
+    fewShotDemoKeyValueList: Optional[str] = None
+    chatBotConfig: Optional[dict] = None
+    sleepInterval: float = 0
+
+
+class TranslationResult(BaseModel):
+    language: str
+    markdown: ocr.MarkdownData
+
+
+class TranslateResult(BaseModel):
+    translationResults: List[TranslationResult]
+
+
+PRIMARY_OPERATIONS: Final[PrimaryOperations] = {
+    "analyzeImages": (
+        ANALYZE_IMAGES_ENDPOINT,
+        AnalyzeImagesRequest,
+        AnalyzeImagesResult,
+    ),
+    "translate": (TRANSLATE_ENDPOINT, TranslateRequest, TranslateResult),
+}

paddlex/inference/serving/schemas/pp_shituv2.py

@@ -90,6 +90,7 @@ class InferRequest(BaseModel):
     recThreshold: Optional[float] = None
     hammingRadius: Optional[float] = None
     topk: Optional[int] = None
+    visualize: Optional[bool] = None
 
 
 class RecResult(BaseModel):

paddlex/inference/serving/schemas/pp_structurev3.py

@@ -22,7 +22,6 @@ from .shared import ocr
 __all__ = [
     "INFER_ENDPOINT",
     "InferRequest",
-    "MarkdownData",
     "LayoutParsingResult",
     "InferResult",
     "PRIMARY_OPERATIONS",
@@ -62,18 +61,12 @@ class InferRequest(ocr.BaseInferRequest):
     useOcrResultsWithTableCells: bool = True
     useE2eWiredTableRecModel: bool = False
     useE2eWirelessTableRecModel: bool = True
-
-
-class MarkdownData(BaseModel):
-    text: str
-    images: Dict[str, str]
-    isStart: bool
-    isEnd: bool
+    visualize: Optional[bool] = None
 
 
 class LayoutParsingResult(BaseModel):
     prunedResult: dict
-    markdown: MarkdownData
+    markdown: ocr.MarkdownData
     outputImages: Optional[Dict[str, str]] = None
     inputImage: Optional[str] = None
 

paddlex/inference/serving/schemas/rotated_object_detection.py

@@ -33,6 +33,7 @@ INFER_ENDPOINT: Final[str] = "/rotated-object-detection"
 class InferRequest(BaseModel):
     image: str
     threshold: Optional[Union[float, Dict[int, float]]] = None
+    visualize: Optional[bool] = None
 
 
 class DetectedObject(BaseModel):

paddlex/inference/serving/schemas/seal_recognition.py

@@ -44,6 +44,7 @@ class InferRequest(ocr.BaseInferRequest):
     sealDetBoxThresh: Optional[float] = None
     sealDetUnclipRatio: Optional[float] = None
     sealRecScoreThresh: Optional[float] = None
+    visualize: Optional[bool] = None
 
 
 class SealRecResult(BaseModel):

paddlex/inference/serving/schemas/semantic_segmentation.py

@@ -32,6 +32,7 @@ INFER_ENDPOINT: Final[str] = "/semantic-segmentation"
 class InferRequest(BaseModel):
     image: str
     targetSize: Optional[Union[int, image_segmentation.Size]] = None
+    visualize: Optional[bool] = None
 
 
 class InferResult(BaseModel):

paddlex/inference/serving/schemas/shared/ocr.py

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Optional
+from typing import Dict, Optional
 
 from pydantic import BaseModel
 from typing_extensions import Literal, TypeAlias
@@ -23,3 +23,10 @@ FileType: TypeAlias = Literal[0, 1]
 class BaseInferRequest(BaseModel):
     file: str
     fileType: Optional[FileType] = None
+
+
+class MarkdownData(BaseModel):
+    text: str
+    isStart: bool
+    isEnd: bool
+    images: Optional[Dict[str, str]] = None

paddlex/inference/serving/schemas/small_object_detection.py

@@ -33,6 +33,7 @@ INFER_ENDPOINT: Final[str] = "/small-object-detection"
 class InferRequest(BaseModel):
     image: str
     threshold: Optional[Union[float, Dict[int, float]]] = None
+    visualize: Optional[bool] = None
 
 
 class DetectedObject(BaseModel):

paddlex/inference/serving/schemas/table_recognition.py

@@ -42,6 +42,7 @@ class InferRequest(ocr.BaseInferRequest):
     textDetUnclipRatio: Optional[float] = None
     textRecScoreThresh: Optional[float] = None
     useOcrResultsWithTableCells: bool = False
+    visualize: Optional[bool] = None
 
 
 class TableRecResult(BaseModel):

paddlex/inference/serving/schemas/table_recognition_v2.py

@@ -47,6 +47,7 @@ class InferRequest(ocr.BaseInferRequest):
     useWirelessTableCellsTransToHtml: bool = False
     useTableOrientationClassify: bool = True
     useOcrResultsWithTableCells: bool = True
+    visualize: Optional[bool] = None
 
 
 class TableRecResult(BaseModel):

paddlex/inference/serving/schemas/ts_anomaly_detection.py

@@ -25,6 +25,7 @@ INFER_ENDPOINT: Final[str] = "/time-series-anomaly-detection"
 
 class InferRequest(BaseModel):
     csv: str
+    visualize: Optional[bool] = None
 
 
 class InferResult(BaseModel):

paddlex/inference/serving/schemas/ts_classification.py

@@ -25,6 +25,7 @@ INFER_ENDPOINT: Final[str] = "/time-series-classification"
 
 class InferRequest(BaseModel):
     csv: str
+    visualize: Optional[bool] = None
 
 
 class InferResult(BaseModel):

paddlex/inference/serving/schemas/ts_forecast.py

@@ -25,6 +25,7 @@ INFER_ENDPOINT: Final[str] = "/time-series-forecasting"
 
 class InferRequest(BaseModel):
     csv: str
+    visualize: Optional[bool] = None
 
 
 class InferResult(BaseModel):

paddlex/inference/serving/schemas/vehicle_attribute_recognition.py

@@ -35,6 +35,7 @@ INFER_ENDPOINT: Final[str] = "/vehicle-attribute-recognition"
 class InferRequest(BaseModel):
     image: str
     detThreshold: Optional[float] = None
+    visualize: Optional[bool] = None
     clsThreshold: Optional[
         Union[float, Dict[Union[Literal["default"], int], float], List[float]]
     ] = None

paddlex/inference/utils/io/__init__.py

@@ -17,6 +17,7 @@ from .readers import (
     AudioReader,
     CSVReader,
     ImageReader,
+    MarkDownReader,
     PDFReader,
     ReaderType,
     VideoReader,

paddlex/inference/utils/io/readers.py

@@ -52,6 +52,8 @@ class ReaderType(enum.Enum):
     TS = 5
     PDF = 6
     YAML = 8
+    MARKDOWN = 9
+    TXT = 10
 
 
 class _BaseReader(object):
@@ -206,6 +208,41 @@ class YAMLReader(_BaseReader):
         return ReaderType.YAML
 
 
+class MarkDownReader(_BaseReader):
+
+    def __init__(self, backend="Markdown", **bk_args):
+        super().__init__(backend, **bk_args)
+
+    def read(self, in_path):
+        return self._backend.read_file(str(in_path))
+
+    def _init_backend(self, bk_type, bk_args):
+        if bk_type == "Markdown":
+            return TXTReaderBackend(**bk_args)
+        else:
+            raise ValueError("Unsupported backend type")
+
+    def get_type(self):
+        return ReaderType.MARKDOWN
+
+
+class TXTReader(_BaseReader):
+    """TXTReader"""
+
+    def __init__(self, backend="txt", **bk_args):
+        super().__init__(backend, **bk_args)
+
+    def read(self, in_path):
+        return self._backend.read_file(str(in_path))
+
+    def _init_backend(self, bk_type, bk_args):
+        if bk_type == "txt":
+            return TXTReaderBackend(**bk_args)
+
+    def get_type(self):
+        return ReaderType.TXT
+
+
 class _BaseReaderBackend(object):
     """_BaseReaderBackend"""
 
@@ -261,6 +298,15 @@ class PDFReaderBackend(_BaseReaderBackend):
             yield img_cv
 
 
+class TXTReaderBackend(_BaseReaderBackend):
+    """TXTReaderBackend"""
+
+    def read_file(self, in_path):
+        with open(in_path, "r") as f:
+            data = f.read()
+        return data
+
+
 class _VideoReaderBackend(_BaseReaderBackend):
     """_VideoReaderBackend"""
 

paddlex/inference/utils/io/writers.py

@@ -54,6 +54,8 @@ class WriterType(enum.Enum):
     XLSX = 6
     CSV = 7
     YAML = 8
+    MARKDOWN = 9
+    TXT = 10
 
 
 class _BaseWriter(object):

paddlex/inference/utils/official_models.py

@@ -359,7 +359,11 @@ PP-LCNet_x1_0_vehicle_attribute_infer.tar",
 PP-OCRv5_server_rec_infer.tar",
     "PP-OCRv5_mobile_rec": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/\
 PP-OCRv5_mobile_rec_infer.tar",
+    "eslav_PP-OCRv5_mobile_rec": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/\
+eslav_PP-OCRv5_mobile_rec_infer.tar",
     "PP-DocBee2-3B": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/PP-DocBee2-3B_infer.tar",
+    "latin_PP-OCRv5_mobile_rec": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/latin_PP-OCRv5_mobile_rec_infer.tar",
+    "korean_PP-OCRv5_mobile_rec": "https://paddle-model-ecology.bj.bcebos.com/paddlex/official_inference_model/paddle3.0.0/korean_PP-OCRv5_mobile_rec_infer.tar",
 }
 
 
@@ -384,6 +388,7 @@ HUGGINGFACE_MODELS = [
     "PicoDet-S_layout_17cls",
     "PicoDet-S_layout_3cls",
     "PP-DocBee2-3B",
+    "PP-Chart2Table",
     "PP-DocBee-2B",
     "PP-DocBee-7B",
     "PP-DocBlockLayout",
@@ -396,8 +401,10 @@ HUGGINGFACE_MODELS = [
     "PP-FormulaNet_plus-M",
     "PP-FormulaNet_plus-S",
     "PP-FormulaNet-S",
+    "PP-LCNet_x0_25_textline_ori",
     "PP-LCNet_x1_0_doc_ori",
     "PP-LCNet_x1_0_table_cls",
+    "PP-LCNet_x1_0_textline_ori",
     "PP-OCRv3_mobile_det",
     "PP-OCRv3_mobile_rec",
     "PP-OCRv3_server_det",

paddlex/modules/anomaly_detection/dataset_checker/dataset_src/convert_dataset.py

@@ -100,9 +100,9 @@ def convert_labelme_dataset(input_dir):
                 if not os.path.exists(img_path):
                     logging.info("%s is not existed, skip this image" % img_path)
                     continue
-                img_name = img_path.split("/")[-1]
+                img_name = osp.basename(img_path)
                 img_file_list.append(f"images/{img_name}")
-                label_img_name = annotated_img_path.split("/")[-1]
+                label_img_name = osp.basename(annotated_img_path)
                 label_file_list.append(f"annotations/{label_img_name}")
 
                 img = np.asarray(cv2.imread(img_path))

paddlex/modules/face_recognition/dataset_checker/dataset_src/utils/visualizer.py

@@ -16,7 +16,7 @@ import numpy as np
 import PIL
 from PIL import ImageDraw, ImageFont
 
-from ......utils.fonts import PINGFANG_FONT_FILE_PATH
+from ......utils.fonts import PINGFANG_FONT
 
 
 def colormap(rgb=False):
@@ -114,7 +114,7 @@ def draw_label(image, label, label_map_dict):
     min_font_size = int(image_size[0] * 0.02)
     max_font_size = int(image_size[0] * 0.05)
     for font_size in range(max_font_size, min_font_size - 1, -1):
-        font = ImageFont.truetype(PINGFANG_FONT_FILE_PATH, font_size, encoding="utf-8")
+        font = ImageFont.truetype(PINGFANG_FONT.path, font_size, encoding="utf-8")
         if tuple(map(int, PIL.__version__.split("."))) <= (10, 0, 0):
             text_width_tmp, text_height_tmp = draw.textsize(
                 label_map_dict[int(label)], font
@@ -127,7 +127,7 @@ def draw_label(image, label, label_map_dict):
         if text_width_tmp <= image_size[0]:
             break
         else:
-            font = ImageFont.truetype(PINGFANG_FONT_FILE_PATH, min_font_size)
+            font = ImageFont.truetype(PINGFANG_FONT.path, min_font_size)
     color_list = colormap(rgb=True)
     color = tuple(color_list[0])
     font_color = tuple(font_colormap(3))

paddlex/modules/formula_recognition/dataset_checker/dataset_src/analyse_dataset.py

@@ -22,7 +22,7 @@ import numpy as np
 
 from .....utils.deps import function_requires_deps, is_dep_available
 from .....utils.file_interface import custom_open
-from .....utils.fonts import PINGFANG_FONT_FILE_PATH
+from .....utils.fonts import PINGFANG_FONT
 from .....utils.logging import warning
 
 if is_dep_available("opencv-contrib-python"):
@@ -128,7 +128,7 @@ def deep_analyse(dataset_path, output, datatype="FormulaRecDataset"):
     if os_system == "windows":
         plt.rcParams["font.sans-serif"] = "FangSong"
     else:
-        font = font_manager.FontProperties(fname=PINGFANG_FONT_FILE_PATH, size=15)
+        font = font_manager.FontProperties(fname=PINGFANG_FONT.path, size=15)
 
     fig, ax = plt.subplots(figsize=(15, 9), dpi=120)
     xlabel_name = "公式长度区间"

paddlex/modules/general_recognition/dataset_checker/dataset_src/analyse_dataset.py

@@ -19,7 +19,7 @@ import numpy as np
 
 from .....utils.deps import function_requires_deps, is_dep_available
 from .....utils.file_interface import custom_open
-from .....utils.fonts import PINGFANG_FONT_FILE_PATH
+from .....utils.fonts import PINGFANG_FONT
 
 if is_dep_available("matplotlib"):
     import matplotlib.pyplot as plt
@@ -52,7 +52,7 @@ def deep_analyse(dataset_path, output, dataset_type="ShiTuRec"):
     if os_system == "windows":
         plt.rcParams["font.sans-serif"] = "FangSong"
     else:
-        font = font_manager.FontProperties(fname=PINGFANG_FONT_FILE_PATH, size=10)
+        font = font_manager.FontProperties(fname=PINGFANG_FONT.path, size=10)
 
     x = np.arange(len(categories))  # 标签位置
     width = 0.35  # 每个条形的宽度

paddlex/modules/general_recognition/dataset_checker/dataset_src/utils/visualizer.py

@@ -16,7 +16,7 @@ import numpy as np
 import PIL
 from PIL import ImageDraw, ImageFont
 
-from ......utils.fonts import PINGFANG_FONT_FILE_PATH
+from ......utils.fonts import PINGFANG_FONT
 
 
 def colormap(rgb=False):
@@ -114,7 +114,7 @@ def draw_label(image, label):
     min_font_size = int(image_size[0] * 0.02)
     max_font_size = int(image_size[0] * 0.05)
     for font_size in range(max_font_size, min_font_size - 1, -1):
-        font = ImageFont.truetype(PINGFANG_FONT_FILE_PATH, font_size, encoding="utf-8")
+        font = ImageFont.truetype(PINGFANG_FONT.path, font_size, encoding="utf-8")
         if tuple(map(int, PIL.__version__.split("."))) <= (10, 0, 0):
             text_width_tmp, text_height_tmp = draw.textsize(label, font)
         else:
@@ -123,7 +123,7 @@ def draw_label(image, label):
         if text_width_tmp <= image_size[0]:
             break
         else:
-            font = ImageFont.truetype(PINGFANG_FONT_FILE_PATH, min_font_size)
+            font = ImageFont.truetype(PINGFANG_FONT.path, min_font_size)
     color_list = colormap(rgb=True)
     color = tuple(color_list[0])
     font_color = tuple(font_colormap(3))

paddlex/modules/image_classification/dataset_checker/dataset_src/analyse_dataset.py

@@ -20,7 +20,7 @@ import numpy as np
 
 from .....utils.deps import function_requires_deps, is_dep_available
 from .....utils.file_interface import custom_open
-from .....utils.fonts import PINGFANG_FONT_FILE_PATH
+from .....utils.fonts import PINGFANG_FONT
 
 if is_dep_available("matplotlib"):
     import matplotlib.pyplot as plt
@@ -68,7 +68,7 @@ def deep_analyse(dataset_path, output):
     if os_system == "windows":
         plt.rcParams["font.sans-serif"] = "FangSong"
     else:
-        font = font_manager.FontProperties(fname=PINGFANG_FONT_FILE_PATH, size=10)
+        font = font_manager.FontProperties(fname=PINGFANG_FONT.path, size=10)
     fig, ax = plt.subplots(figsize=(max(8, int(len(classes) / 5)), 5), dpi=300)
     ax.bar(x, cnts_train_sorted, width=0.5, label="train")
     ax.bar(x + width, cnts_val_sorted, width=0.5, label="val")

paddlex/modules/image_classification/dataset_checker/dataset_src/utils/visualizer.py

@@ -16,7 +16,7 @@ import numpy as np
 import PIL
 from PIL import ImageDraw, ImageFont
 
-from ......utils.fonts import PINGFANG_FONT_FILE_PATH
+from ......utils.fonts import PINGFANG_FONT
 
 
 def colormap(rgb=False):
@@ -114,7 +114,7 @@ def draw_label(image, label, label_map_dict):
     min_font_size = int(image_size[0] * 0.02)
     max_font_size = int(image_size[0] * 0.05)
     for font_size in range(max_font_size, min_font_size - 1, -1):
-        font = ImageFont.truetype(PINGFANG_FONT_FILE_PATH, font_size, encoding="utf-8")
+        font = ImageFont.truetype(PINGFANG_FONT.path, font_size, encoding="utf-8")
         if tuple(map(int, PIL.__version__.split("."))) <= (10, 0, 0):
             text_width_tmp, text_height_tmp = draw.textsize(
                 label_map_dict[int(label)], font
@@ -127,7 +127,7 @@ def draw_label(image, label, label_map_dict):
         if text_width_tmp <= image_size[0]:
             break
         else:
-            font = ImageFont.truetype(PINGFANG_FONT_FILE_PATH, min_font_size)
+            font = ImageFont.truetype(PINGFANG_FONT.path, min_font_size)
     color_list = colormap(rgb=True)
     color = tuple(color_list[0])
     font_color = tuple(font_colormap(3))

paddlex/modules/instance_segmentation/dataset_checker/dataset_src/analyse_dataset.py

@@ -20,7 +20,7 @@ from collections import defaultdict
 import numpy as np
 
 from .....utils.deps import function_requires_deps, is_dep_available
-from .....utils.fonts import PINGFANG_FONT_FILE_PATH
+from .....utils.fonts import PINGFANG_FONT
 
 if is_dep_available("matplotlib"):
     import matplotlib.pyplot as plt
@@ -64,7 +64,7 @@ def deep_analyse(dataset_dir, output):
     if os_system == "windows":
         plt.rcParams["font.sans-serif"] = "FangSong"
     else:
-        font = font_manager.FontProperties(fname=PINGFANG_FONT_FILE_PATH)
+        font = font_manager.FontProperties(fname=PINGFANG_FONT.path)
     fig, ax = plt.subplots(figsize=(max(8, int(len(classes) / 5)), 5), dpi=120)
     ax.bar(x, cnts_train_sorted, width=0.5, label="train")
     ax.bar(x + width, cnts_val_sorted, width=0.5, label="val")

paddlex/modules/instance_segmentation/dataset_checker/dataset_src/utils/visualizer.py

@@ -18,7 +18,7 @@ from PIL import Image, ImageDraw, ImageFont
 
 from ......utils import logging
 from ......utils.deps import function_requires_deps, is_dep_available
-from ......utils.fonts import PINGFANG_FONT_FILE_PATH
+from ......utils.fonts import PINGFANG_FONT
 
 if is_dep_available("pycocotools"):
     from pycocotools.coco import COCO
@@ -124,7 +124,7 @@ def draw_bbox(image, coco_info: "COCO", img_id):
         font_size = int(0.024 * int(image_info["width"])) + 2
     except:
         font_size = 12
-    font = ImageFont.truetype(PINGFANG_FONT_FILE_PATH, font_size, encoding="utf-8")
+    font = ImageFont.truetype(PINGFANG_FONT.path, font_size, encoding="utf-8")
 
     image = image.convert("RGB")
     draw = ImageDraw.Draw(image)

paddlex/modules/m_3d_bev_detection/dataset_checker/dataset_src/analyse_dataset.py

@@ -21,7 +21,7 @@ from collections import defaultdict
 import numpy as np
 
 from paddlex.utils.deps import function_requires_deps, is_dep_available
-from paddlex.utils.fonts import PINGFANG_FONT_FILE_PATH
+from paddlex.utils.fonts import PINGFANG_FONT
 
 if is_dep_available("matplotlib"):
     import matplotlib.pyplot as plt
@@ -88,7 +88,7 @@ def deep_analyse(dataset_dir, output):
     if os_system == "windows":
         plt.rcParams["font.sans-serif"] = "FangSong"
     else:
-        font = font_manager.FontProperties(fname=PINGFANG_FONT_FILE_PATH)
+        font = font_manager.FontProperties(fname=PINGFANG_FONT.path)
     fig, ax = plt.subplots(figsize=(max(8, int(len(classes) / 5)), 5), dpi=120)
     ax.bar(x, cnts_train_sorted, width=0.5, label="train")
     ax.bar(x + width, cnts_val_sorted, width=0.5, label="val")