No public description

PiperOrigin-RevId: 604470925

Author: tensorflower-gardener

official/projects/waste_identification_ml/docker_solution/prediction_pipeline/mask_bbox_saver.py

@@ -26,7 +26,7 @@
 
 import os
 import sys
-from typing import Any
+from typing import Any, Dict
 import cv2
 import numpy as np
 
@@ -35,11 +35,11 @@
 
 
 def save_bbox_masks_labels(
-    result: dict[Any, np.ndarray],
+    result: Dict[Any, np.ndarray],
     image: np.ndarray,
     file_name: str,
     folder: str,
-    category_index: dict[int, dict[str, str]],
+    category_index: Dict[int, Dict[str, str]],
     threshold: float,
 ) -> None:
   """Saves an image with visualized bounding boxes, labels, and masks.
@@ -86,7 +86,7 @@ def save_bbox_masks_labels(
 
 
 def save_binary_masks(
-    result: dict[Any, np.ndarray], file_name: str, folder: str
+    result: Dict[Any, np.ndarray], file_name: str, folder: str
 ) -> None:
   """Saves binary masks generated from object detection results.
 

official/projects/waste_identification_ml/model_inference/color_and_property_extractor.py

@@ -14,7 +14,7 @@
 
 """Extract properties from each object mask and detect its color."""
 
-from typing import Annotated, Literal, TypeVar, Union
+from typing import Dict, List, Tuple, TypeVar, Union
 
 import numpy as np
 import numpy.typing as npt
@@ -30,7 +30,8 @@
 # Those values could be in different scales like
 # RGB ([0.0,255.0], [0.0,255.0], [0.0 to 255.0])
 # LAB ([0.0,100], [-128,127], [-128,127])
-NColor = Annotated[npt.NDArray[DType], Literal[3]][np.float64]
+# NColor = Annotated[npt.NDArray[DType], Literal[3]][np.float64]
+NColor = np.ndarray
 
 
 PROPERTIES = [
@@ -108,11 +109,11 @@
 
 
 def extract_properties_and_object_masks(
-    final_result: dict[str, np.ndarray],
+    final_result: Dict[str, np.ndarray],
     height: int,
     width: int,
     original_image: np.ndarray,
-) -> tuple[list[pd.DataFrame], list[np.ndarray]]:
+) -> Tuple[List[pd.DataFrame], List[np.ndarray]]:
   """Extract specific properties from given detection masks.
 
   Properties that will be computed includes the area of the masks, bbox
@@ -163,7 +164,7 @@ def extract_properties_and_object_masks(
 
 def find_dominant_color(
     image: np.ndarray, black_threshold: int = 50
-) -> tuple[Union[int, str], Union[int, str], Union[int, str]]:
+) -> Tuple[Union[int, str], Union[int, str], Union[int, str]]:
   """Determines the dominant color in a given image.
 
   The function performs the following steps:
@@ -212,7 +213,7 @@ def color_difference(color1: int, color2: int) -> Union[float, int]:
   return (color1 - color2) ** 2
 
 
-def est_color(requested_color: tuple[int, int, int]) -> str:
+def est_color(requested_color: Tuple[int, int, int]) -> str:
   """Estimates the closest named color for a given RGB color.
 
   The function uses the Euclidean distance in the RGB space to find the closest
@@ -238,7 +239,7 @@ def est_color(requested_color: tuple[int, int, int]) -> str:
   return min_colors[min(min_colors.keys())]
 
 
-def get_color_name(rgb_color: tuple[int, int, int]) -> str | None:
+def get_color_name(rgb_color: Tuple[int, int, int]) -> str | None:
   """Retrieves the name of a given RGB color.
 
   If the RGB color exactly matches one of the CSS3 predefined colors, it returns
@@ -265,7 +266,7 @@ def get_color_name(rgb_color: tuple[int, int, int]) -> str | None:
     return None
 
 
-def rgb_int_to_lab(rgb_int_color: tuple[int, int, int]) -> NColor:
+def rgb_int_to_lab(rgb_int_color: Tuple[int, int, int]) -> NColor:
   """Convert RGB color to LAB color space.
 
   Args:
@@ -280,7 +281,7 @@ def rgb_int_to_lab(rgb_int_color: tuple[int, int, int]) -> NColor:
 
 
 def color_distance(
-    a: tuple[int, int, int], b: tuple[int, int, int]
+    a: Tuple[int, int, int], b: Tuple[int, int, int]
 ) -> np.ndarray:
   """The color distance following the ciede2000 formula.
 
@@ -297,8 +298,8 @@ def color_distance(
 
 
 def build_color_lab_list(
-    generic_colors: list[tuple[str, str]]
-) -> tuple[npt.NDArray[np.str_], list[NColor]]:
+    generic_colors: List[Tuple[str, str]]
+) -> Tuple[npt.NDArray[np.str_], List[NColor]]:
   """Get Simple colors names and lab values.
 
   Args:
@@ -326,9 +327,9 @@ def build_color_lab_list(
 
 
 def get_generic_color_name(
-    rgb_colors: list[tuple[int, int, int]],
-    generic_colors: list[tuple[str, str]] | None = None,
-) -> list[str]:
+    rgb_colors: List[Tuple[int, int, int]],
+    generic_colors: List[Tuple[str, str]] | None = None,
+) -> List[str]:
   """Retrieves generic names of given RGB colors.
 
   Estimates the closest matching color name.

official/projects/waste_identification_ml/model_inference/postprocessing.py

@@ -28,7 +28,7 @@
 object should not have more than a single bounding box. 
 """
 import copy
-from typing import Any, Optional, TypedDict
+from typing import Any, Optional, TypedDict, Dict, Tuple, List
 import numpy as np
 import tensorflow as tf, tf_keras
 
@@ -128,8 +128,8 @@ def reframe_box_masks_to_image_masks_default():
 
 
 def reframing_masks(
-    results: dict[str, np.ndarray], height: int, width: int
-) -> dict[str, np.ndarray]:
+    results: Dict[str, np.ndarray], height: int, width: int
+) -> Dict[str, np.ndarray]:
   """Processes the output from Mask RCNN model to create a full size mask.
 
   Args:
@@ -155,7 +155,7 @@ def reframing_masks(
 
 
 def find_id_by_name(
-    dictionary: dict[int, ItemDict], name: str
+    dictionary: Dict[int, ItemDict], name: str
 ) -> Optional[int]:
   """Finds the id of a dictionary given its value.
 
@@ -178,9 +178,9 @@ def find_id_by_name(
 
 
 def combine_bounding_boxes(
-    box1: list[float],
-    box2: list[float],
-) -> list[float]:
+    box1: List[float],
+    box2: List[float],
+) -> List[float]:
   """Combines two bounding boxes.
 
   Args:
@@ -207,9 +207,9 @@ def calculate_combined_scores_boxes_classes(
     j: int,
     results_1: DetectionResult,
     results_2: DetectionResult,
-    category_indices: list[list[Any]],
-    category_index_combined: dict[int, ItemDict],
-) -> tuple[Any, list[float], Any, Optional[int]]:
+    category_indices: List[List[Any]],
+    category_index_combined: Dict[int, ItemDict],
+) -> Tuple[Any, List[float], Any, Optional[int]]:
   """Calculate combined scores, boxes, and classes for matched masks.
 
   Args:
@@ -249,9 +249,9 @@ def calculate_combined_scores_boxes_classes(
 def calculate_single_result(
     index: int,
     result: DetectionResult,
-    category_indices: list[list[Any]],
+    category_indices: List[List[Any]],
     flag: Any | str,
-) -> tuple[float, tuple[float, float, float, float], str]:
+) -> Tuple[float, Tuple[float, float, float, float], str]:
   """Calculate scores, boxes, and classes for non-matched masks.
 
   Args:
@@ -279,7 +279,7 @@ def calculate_single_result(
 
 def calculate_iou(
     mask1: np.ndarray, mask2: np.ndarray
-) -> tuple[float, np.ndarray]:
+) -> Tuple[float, np.ndarray]:
   """Calculates the intersection over union (IoU) score for two masks.
 
   Args:
@@ -305,11 +305,11 @@ def find_similar_masks(
     results_2: DetectionResult,
     num_detections: int,
     min_score_thresh: float,
-    category_indices: list[list[Any]],
-    category_index_combined: dict[int, ItemDict],
+    category_indices: List[List[Any]],
+    category_index_combined: Dict[int, ItemDict],
     area_threshold: float,
     iou_threshold: float = 0.8,
-) -> dict[str, np.ndarray]:
+) -> Dict[str, np.ndarray]:
   """Aligns the masks of the detections in `results_1` and `results_2`.
 
   Args:
@@ -437,10 +437,10 @@ def find_similar_masks(
 
 
 def filter_bounding_boxes(
-    bounding_boxes: list[tuple[int, int, int, int]],
+    bounding_boxes: List[Tuple[int, int, int, int]],
     iou_threshold: float = 0.5,
     area_ratio_threshold: float = 0.8,
-) -> tuple[list[tuple[int, int, int, int]], list[int]]:
+) -> Tuple[List[Tuple[int, int, int, int]], List[int]]:
   """Filters overlapping bounding boxes based on IoU and area ratio criteria.
 
   This function filters out overlapping bounding boxes from a given list based