ultralytics 8.3.78__tar.gz → 8.3.80__tar.gz

{ultralytics-8.3.78/ultralytics.egg-info → ultralytics-8.3.80}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: ultralytics
-Version: 8.3.78
+Version: 8.3.80
 Summary: Ultralytics YOLO 🚀 for SOTA object detection, multi-object tracking, instance segmentation, pose estimation and image classification.
 Author-email: Glenn Jocher <glenn.jocher@ultralytics.com>, Jing Qiu <jing.qiu@ultralytics.com>
 Maintainer-email: Ultralytics <hello@ultralytics.com>
@@ -63,7 +63,7 @@ Provides-Extra: export
 Requires-Dist: onnx>=1.12.0; extra == "export"
 Requires-Dist: coremltools>=7.0; (platform_system != "Windows" and python_version <= "3.11") and extra == "export"
 Requires-Dist: scikit-learn>=1.3.2; (platform_system != "Windows" and python_version <= "3.11") and extra == "export"
-Requires-Dist: openvino>=2024.0.0; extra == "export"
+Requires-Dist: openvino<2025.0.0,>=2024.0.0; extra == "export"
 Requires-Dist: tensorflow>=2.0.0; extra == "export"
 Requires-Dist: tensorflowjs>=3.9.0; extra == "export"
 Requires-Dist: tensorstore>=0.1.63; (platform_machine == "aarch64" and python_version >= "3.9") and extra == "export"

{ultralytics-8.3.78 → ultralytics-8.3.80}/pyproject.toml

@@ -97,7 +97,7 @@ export = [
     "onnx>=1.12.0", # ONNX export
     "coremltools>=7.0; platform_system != 'Windows' and python_version <= '3.11'", # CoreML supported on macOS and Linux
     "scikit-learn>=1.3.2; platform_system != 'Windows' and python_version <= '3.11'", # CoreML k-means quantization
-    "openvino>=2024.0.0", # OpenVINO export
+    "openvino>=2024.0.0,<2025.0.0", # OpenVINO export
     "tensorflow>=2.0.0", # TF bug https://github.com/ultralytics/ultralytics/issues/5161
     "tensorflowjs>=3.9.0", # TF.js export, automatically installs tensorflow
     "tensorstore>=0.1.63; platform_machine == 'aarch64' and python_version >= '3.9'", # for TF Raspberry Pi exports

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/__init__.py

@@ -1,6 +1,6 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-__version__ = "8.3.78"
+__version__ = "8.3.80"
 
 import os
 

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/data/annotator.py

@@ -58,15 +58,13 @@ def auto_annotate(
 
     for result in det_results:
         class_ids = result.boxes.cls.int().tolist()  # noqa
-        if len(class_ids):
+        if class_ids:
             boxes = result.boxes.xyxy  # Boxes object for bbox outputs
             sam_results = sam_model(result.orig_img, bboxes=boxes, verbose=False, save=False, device=device)
             segments = sam_results[0].masks.xyn  # noqa
 
             with open(f"{Path(output_dir) / Path(result.path).stem}.txt", "w") as f:
-                for i in range(len(segments)):
-                    s = segments[i]
-                    if len(s) == 0:
-                        continue
-                    segment = map(str, segments[i].reshape(-1).tolist())
-                    f.write(f"{class_ids[i]} " + " ".join(segment) + "\n")
+                for i, s in enumerate(segments):
+                    if s.any():
+                        segment = map(str, s.reshape(-1).tolist())
+                        f.write(f"{class_ids[i]} " + " ".join(segment) + "\n")

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/data/augment.py

@@ -1364,17 +1364,26 @@ class RandomHSV:
             >>> hsv_augmenter(labels)
             >>> augmented_img = labels["img"]
         """
-        img = labels["img"]
         if self.hgain or self.sgain or self.vgain:
-            r = np.random.uniform(-1, 1, 3) * [self.hgain, self.sgain, self.vgain] + 1  # random gains
-            hue, sat, val = cv2.split(cv2.cvtColor(img, cv2.COLOR_BGR2HSV))
+            img = labels["img"]
             dtype = img.dtype  # uint8
 
+            # Original implementation (bug) from ultralytics<=8.3.78
+            # r = np.random.uniform(-1, 1, 3) * [self.hgain, self.sgain, self.vgain] + 1  # random gains
+            # x = np.arange(0, 256, dtype=r.dtype)
+            # lut_hue = ((x * r[0]) % 180).astype(dtype)
+            # lut_sat = np.clip(x * r[1], 0, 255).astype(dtype)
+            # lut_val = np.clip(x * r[2], 0, 255).astype(dtype)
+
+            # Fixed implementation in https://github.com/ultralytics/ultralytics/pull/19311
+            r = np.random.uniform(-1, 1, 3) * (self.hgain, self.sgain, self.vgain) * (180, 255, 255)  # random gains
             x = np.arange(0, 256, dtype=r.dtype)
-            lut_hue = ((x * r[0]) % 180).astype(dtype)
-            lut_sat = np.clip(x * r[1], 0, 255).astype(dtype)
-            lut_val = np.clip(x * r[2], 0, 255).astype(dtype)
+            lut_hue = ((x + r[0]) % 180).astype(dtype)
+            lut_sat = np.clip(x + r[1], 0, 255).astype(dtype)
+            lut_val = np.clip(x + r[2], 0, 255).astype(dtype)
+            lut_sat[0] = 0  # prevent pure white changing color
 
+            hue, sat, val = cv2.split(cv2.cvtColor(img, cv2.COLOR_BGR2HSV))
             im_hsv = cv2.merge((cv2.LUT(hue, lut_hue), cv2.LUT(sat, lut_sat), cv2.LUT(val, lut_val)))
             cv2.cvtColor(im_hsv, cv2.COLOR_HSV2BGR, dst=img)  # no return needed
         return labels
@@ -1484,7 +1493,7 @@ class LetterBox:
     Attributes:
         new_shape (tuple): Target shape (height, width) for resizing.
         auto (bool): Whether to use minimum rectangle.
-        scaleFill (bool): Whether to stretch the image to new_shape.
+        scale_fill (bool): Whether to stretch the image to new_shape.
         scaleup (bool): Whether to allow scaling up. If False, only scale down.
         stride (int): Stride for rounding padding.
         center (bool): Whether to center the image or align to top-left.
@@ -1499,7 +1508,7 @@ class LetterBox:
         >>> updated_instances = result["instances"]
     """
 
-    def __init__(self, new_shape=(640, 640), auto=False, scaleFill=False, scaleup=True, center=True, stride=32):
+    def __init__(self, new_shape=(640, 640), auto=False, scale_fill=False, scaleup=True, center=True, stride=32):
         """
         Initialize LetterBox object for resizing and padding images.
 
@@ -1509,7 +1518,7 @@ class LetterBox:
         Args:
             new_shape (Tuple[int, int]): Target size (height, width) for the resized image.
             auto (bool): If True, use minimum rectangle to resize. If False, use new_shape directly.
-            scaleFill (bool): If True, stretch the image to new_shape without padding.
+            scale_fill (bool): If True, stretch the image to new_shape without padding.
             scaleup (bool): If True, allow scaling up. If False, only scale down.
             center (bool): If True, center the placed image. If False, place image in top-left corner.
             stride (int): Stride of the model (e.g., 32 for YOLOv5).
@@ -1517,17 +1526,17 @@ class LetterBox:
         Attributes:
             new_shape (Tuple[int, int]): Target size for the resized image.
             auto (bool): Flag for using minimum rectangle resizing.
-            scaleFill (bool): Flag for stretching image without padding.
+            scale_fill (bool): Flag for stretching image without padding.
             scaleup (bool): Flag for allowing upscaling.
             stride (int): Stride value for ensuring image size is divisible by stride.
 
         Examples:
-            >>> letterbox = LetterBox(new_shape=(640, 640), auto=False, scaleFill=False, scaleup=True, stride=32)
+            >>> letterbox = LetterBox(new_shape=(640, 640), auto=False, scale_fill=False, scaleup=True, stride=32)
             >>> resized_img = letterbox(original_img)
         """
         self.new_shape = new_shape
         self.auto = auto
-        self.scaleFill = scaleFill
+        self.scale_fill = scale_fill
         self.scaleup = scaleup
         self.stride = stride
         self.center = center  # Put the image in the middle or top-left
@@ -1573,7 +1582,7 @@ class LetterBox:
         dw, dh = new_shape[1] - new_unpad[0], new_shape[0] - new_unpad[1]  # wh padding
         if self.auto:  # minimum rectangle
             dw, dh = np.mod(dw, self.stride), np.mod(dh, self.stride)  # wh padding
-        elif self.scaleFill:  # stretch
+        elif self.scale_fill:  # stretch
             dw, dh = 0.0, 0.0
             new_unpad = (new_shape[1], new_shape[0])
             ratio = new_shape[1] / shape[1], new_shape[0] / shape[0]  # width, height ratios

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/data/converter.py

@@ -274,13 +274,13 @@ def convert_coco(
         # Create image dict
         images = {f"{x['id']:d}": x for x in data["images"]}
         # Create image-annotations dict
-        imgToAnns = defaultdict(list)
+        annotations = defaultdict(list)
         for ann in data["annotations"]:
-            imgToAnns[ann["image_id"]].append(ann)
+            annotations[ann["image_id"]].append(ann)
 
         image_txt = []
         # Write labels file
-        for img_id, anns in TQDM(imgToAnns.items(), desc=f"Annotations {json_file}"):
+        for img_id, anns in TQDM(annotations.items(), desc=f"Annotations {json_file}"):
             img = images[f"{img_id:d}"]
             h, w = img["height"], img["width"]
             f = str(Path(img["coco_url"]).relative_to("http://images.cocodataset.org")) if lvis else img["file_name"]

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/engine/exporter.py

@@ -170,6 +170,7 @@ def try_export(inner_func):
     def outer_func(*args, **kwargs):
         """Export a model."""
         prefix = inner_args["prefix"]
+        dt = 0.0
         try:
             with Profile() as dt:
                 f, model = inner_func(*args, **kwargs)
@@ -569,7 +570,7 @@ class Exporter:
     @try_export
     def export_openvino(self, prefix=colorstr("OpenVINO:")):
         """YOLO OpenVINO export."""
-        check_requirements("openvino>=2024.5.0")
+        check_requirements("openvino>=2024.0.0,<2025.0.0")
         import openvino as ov
 
         LOGGER.info(f"\n{prefix} starting export with openvino {ov.__version__}...")
@@ -591,7 +592,7 @@ class Exporter:
             if self.model.task != "classify":
                 ov_model.set_rt_info("fit_to_window_letterbox", ["model_info", "resize_type"])
 
-            ov.runtime.save_model(ov_model, file, compress_to_fp16=self.args.half)
+            ov.save_model(ov_model, file, compress_to_fp16=self.args.half)
             yaml_save(Path(file).parent / "metadata.yaml", self.metadata)  # add metadata.yaml
 
         if self.args.int8:

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/engine/trainer.py

@@ -567,6 +567,10 @@ class BaseTrainer:
         except Exception as e:
             raise RuntimeError(emojis(f"Dataset '{clean_url(self.args.data)}' error ❌ {e}")) from e
         self.data = data
+        if self.args.single_cls:
+            LOGGER.info("Overriding class names with single class.")
+            self.data["names"] = {0: "item"}
+            self.data["nc"] = 1
         return data["train"], data.get("val") or data.get("test")
 
     def setup_model(self):

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/models/nas/model.py

@@ -16,6 +16,7 @@ from pathlib import Path
 import torch
 
 from ultralytics.engine.model import Model
+from ultralytics.utils import DEFAULT_CFG_DICT
 from ultralytics.utils.downloads import attempt_download_asset
 from ultralytics.utils.torch_utils import model_info
 
@@ -77,6 +78,7 @@ class NAS(Model):
         self.model.yaml = {}  # for info()
         self.model.pt_path = weights  # for export()
         self.model.task = "detect"  # for export()
+        self.model.args = {**DEFAULT_CFG_DICT, **self.overrides}  # for export()
 
     def info(self, detailed=False, verbose=True):
         """

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/models/rtdetr/predict.py

@@ -72,7 +72,7 @@ class RTDETRPredictor(BasePredictor):
     def pre_transform(self, im):
         """
         Pre-transforms the input images before feeding them into the model for inference. The input images are
-        letterboxed to ensure a square aspect ratio and scale-filled. The size must be square(640) and scaleFilled.
+        letterboxed to ensure a square aspect ratio and scale-filled. The size must be square(640) and scale_filled.
 
         Args:
             im (list[np.ndarray] |torch.Tensor): Input images of shape (N,3,h,w) for tensor, [(h,w,3) x N] for list.
@@ -80,5 +80,5 @@ class RTDETRPredictor(BasePredictor):
         Returns:
             (list): List of pre-transformed images ready for model inference.
         """
-        letterbox = LetterBox(self.imgsz, auto=False, scaleFill=True)
+        letterbox = LetterBox(self.imgsz, auto=False, scale_fill=True)
         return [letterbox(image=x) for x in im]

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/models/rtdetr/val.py

@@ -34,7 +34,7 @@ class RTDETRDataset(YOLODataset):
             hyp.mixup = hyp.mixup if self.augment and not self.rect else 0.0
             transforms = v8_transforms(self, self.imgsz, hyp, stretch=True)
         else:
-            # transforms = Compose([LetterBox(new_shape=(self.imgsz, self.imgsz), auto=False, scaleFill=True)])
+            # transforms = Compose([LetterBox(new_shape=(self.imgsz, self.imgsz), auto=False, scale_fill=True)])
             transforms = Compose([])
         transforms.append(
             Format(

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/models/yolo/detect/val.py

@@ -186,7 +186,7 @@ class DetectionValidator(BaseValidator):
         self.nt_per_class = np.bincount(stats["target_cls"].astype(int), minlength=self.nc)
         self.nt_per_image = np.bincount(stats["target_img"].astype(int), minlength=self.nc)
         stats.pop("target_img", None)
-        if len(stats) and stats["tp"].any():
+        if len(stats):
             self.metrics.process(**stats)
         return self.metrics.results_dict
 

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/nn/autobackend.py

@@ -244,7 +244,7 @@ class AutoBackend(nn.Module):
         # OpenVINO
         elif xml:
             LOGGER.info(f"Loading {w} for OpenVINO inference...")
-            check_requirements("openvino>=2024.0.0")
+            check_requirements("openvino>=2024.0.0,<2025.0.0")
             import openvino as ov
 
             core = ov.Core()
@@ -600,7 +600,7 @@ class AutoBackend(nn.Module):
                     results[userdata] = request.results
 
                 # Create AsyncInferQueue, set the callback and start asynchronous inference for each input image
-                async_queue = self.ov.runtime.AsyncInferQueue(self.ov_compiled_model)
+                async_queue = self.ov.AsyncInferQueue(self.ov_compiled_model)
                 async_queue.set_callback(callback)
                 for i in range(n):
                     # Start async inference with userdata=i to specify the position in results list

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/nn/modules/block.py

@@ -1227,8 +1227,8 @@ class AAttn(nn.Module):
             v = v.reshape(B // self.area, N * self.area, C)
             B, N, _ = x.shape
 
-        x = x.reshape(B, H, W, C).permute(0, 3, 1, 2)
-        v = v.reshape(B, H, W, C).permute(0, 3, 1, 2)
+        x = x.reshape(B, H, W, C).permute(0, 3, 1, 2).contiguous()
+        v = v.reshape(B, H, W, C).permute(0, 3, 1, 2).contiguous()
 
         x = x + self.pe(v)
         return self.proj(x)

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/utils/callbacks/comet.py

@@ -177,7 +177,7 @@ def _format_ground_truth_annotations_for_detection(img_idx, image_path, batch, c
     return {"name": "ground_truth", "data": data}
 
 
-def _format_prediction_annotations_for_detection(image_path, metadata, class_label_map=None):
+def _format_prediction_annotations_for_detection(image_path, metadata, class_label_map=None, class_map=None):
     """Format YOLO predictions for object detection visualization."""
     stem = image_path.stem
     image_id = int(stem) if stem.isnumeric() else stem
@@ -187,26 +187,32 @@ def _format_prediction_annotations_for_detection(image_path, metadata, class_lab
         LOGGER.debug(f"COMET WARNING: Image: {image_path} has no bounding boxes predictions")
         return None
 
+    label_index_offset = 0
+    if class_map is not None:
+        # offset to align indices of class labels (starting from zero)
+        # with prediction's category ID indices (can start from one)
+        label_index_offset = sorted(class_map)[0]
+
     data = []
     for prediction in predictions:
         boxes = prediction["bbox"]
         score = _scale_confidence_score(prediction["score"])
         cls_label = prediction["category_id"]
         if class_label_map:
-            cls_label = str(class_label_map[cls_label])
+            cls_label = str(class_label_map[cls_label - label_index_offset])
 
         data.append({"boxes": [boxes], "label": cls_label, "score": score})
 
     return {"name": "prediction", "data": data}
 
 
-def _fetch_annotations(img_idx, image_path, batch, prediction_metadata_map, class_label_map):
+def _fetch_annotations(img_idx, image_path, batch, prediction_metadata_map, class_label_map, class_map):
     """Join the ground truth and prediction annotations if they exist."""
     ground_truth_annotations = _format_ground_truth_annotations_for_detection(
         img_idx, image_path, batch, class_label_map
     )
     prediction_annotations = _format_prediction_annotations_for_detection(
-        image_path, prediction_metadata_map, class_label_map
+        image_path, prediction_metadata_map, class_label_map, class_map
     )
 
     annotations = [
@@ -260,6 +266,7 @@ def _log_image_predictions(experiment, validator, curr_step):
     predictions_metadata_map = _create_prediction_metadata_map(jdict)
     dataloader = validator.dataloader
     class_label_map = validator.names
+    class_map = getattr(validator, "class_map", None)
 
     batch_logging_interval = _get_eval_batch_logging_interval()
     max_image_predictions = _get_max_image_predictions_to_log()
@@ -280,6 +287,7 @@ def _log_image_predictions(experiment, validator, curr_step):
                 batch,
                 predictions_metadata_map,
                 class_label_map,
+                class_map=class_map,
             )
             _log_images(
                 experiment,

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/utils/metrics.py

@@ -604,7 +604,7 @@ def ap_per_class(
             if j == 0:
                 prec_values.append(np.interp(x, mrec, mpre))  # precision at mAP@0.5
 
-    prec_values = np.array(prec_values)  # (nc, 1000)
+    prec_values = np.array(prec_values) if prec_values else np.zeros((1, 1000))  # (nc, 1000)
 
     # Compute F1 (harmonic mean of precision and recall)
     f1_curve = 2 * p_curve * r_curve / (p_curve + r_curve + eps)

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics/utils/ops.py

@@ -799,10 +799,11 @@ def regularize_rboxes(rboxes):
         (torch.Tensor): The regularized boxes.
     """
     x, y, w, h, t = rboxes.unbind(dim=-1)
-    # Swap edge and angle if h >= w
-    w_ = torch.where(w > h, w, h)
-    h_ = torch.where(w > h, h, w)
-    t = torch.where(w > h, t, t + math.pi / 2) % math.pi
+    # Swap edge if t >= pi/2 while not being symmetrically opposite
+    swap = t % math.pi >= math.pi / 2
+    w_ = torch.where(swap, h, w)
+    h_ = torch.where(swap, w, h)
+    t = t % (math.pi / 2)
     return torch.stack([x, y, w_, h_, t], dim=-1)  # regularized boxes
 
 

{ultralytics-8.3.78 → ultralytics-8.3.80/ultralytics.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: ultralytics
-Version: 8.3.78
+Version: 8.3.80
 Summary: Ultralytics YOLO 🚀 for SOTA object detection, multi-object tracking, instance segmentation, pose estimation and image classification.
 Author-email: Glenn Jocher <glenn.jocher@ultralytics.com>, Jing Qiu <jing.qiu@ultralytics.com>
 Maintainer-email: Ultralytics <hello@ultralytics.com>
@@ -63,7 +63,7 @@ Provides-Extra: export
 Requires-Dist: onnx>=1.12.0; extra == "export"
 Requires-Dist: coremltools>=7.0; (platform_system != "Windows" and python_version <= "3.11") and extra == "export"
 Requires-Dist: scikit-learn>=1.3.2; (platform_system != "Windows" and python_version <= "3.11") and extra == "export"
-Requires-Dist: openvino>=2024.0.0; extra == "export"
+Requires-Dist: openvino<2025.0.0,>=2024.0.0; extra == "export"
 Requires-Dist: tensorflow>=2.0.0; extra == "export"
 Requires-Dist: tensorflowjs>=3.9.0; extra == "export"
 Requires-Dist: tensorstore>=0.1.63; (platform_machine == "aarch64" and python_version >= "3.9") and extra == "export"

{ultralytics-8.3.78 → ultralytics-8.3.80}/ultralytics.egg-info/requires.txt

@@ -31,7 +31,7 @@ mkdocs-macros-plugin>=1.0.5
 
 [export]
 onnx>=1.12.0
-openvino>=2024.0.0
+openvino<2025.0.0,>=2024.0.0
 tensorflow>=2.0.0
 tensorflowjs>=3.9.0
 keras