dgenerate-ultralytics-headless 8.3.236__py3-none-any.whl → 8.3.239__py3-none-any.whl

ultralytics/cfg/datasets/VOC.yaml

@@ -1,7 +1,7 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
 # PASCAL VOC dataset http://host.robots.ox.ac.uk/pascal/VOC by University of Oxford
-# Documentation: # Documentation: https://docs.ultralytics.com/datasets/detect/voc/
+# Documentation: https://docs.ultralytics.com/datasets/detect/voc/
 # Example usage: yolo train data=VOC.yaml
 # parent
 # ├── ultralytics
@@ -59,22 +59,21 @@ download: |
           x, y, w, h = (box[0] + box[1]) / 2.0 - 1, (box[2] + box[3]) / 2.0 - 1, box[1] - box[0], box[3] - box[2]
           return x * dw, y * dh, w * dw, h * dh
 
-      in_file = open(path / f"VOC{year}/Annotations/{image_id}.xml")
-      out_file = open(lb_path, "w")
-      tree = ET.parse(in_file)
-      root = tree.getroot()
-      size = root.find("size")
-      w = int(size.find("width").text)
-      h = int(size.find("height").text)
+      with open(path / f"VOC{year}/Annotations/{image_id}.xml") as in_file, open(lb_path, "w", encoding="utf-8") as out_file:
+          tree = ET.parse(in_file)
+          root = tree.getroot()
+          size = root.find("size")
+          w = int(size.find("width").text)
+          h = int(size.find("height").text)
 
-      names = list(yaml["names"].values())  # names list
-      for obj in root.iter("object"):
-          cls = obj.find("name").text
-          if cls in names and int(obj.find("difficult").text) != 1:
-              xmlbox = obj.find("bndbox")
-              bb = convert_box((w, h), [float(xmlbox.find(x).text) for x in ("xmin", "xmax", "ymin", "ymax")])
-              cls_id = names.index(cls)  # class id
-              out_file.write(" ".join(str(a) for a in (cls_id, *bb)) + "\n")
+          names = list(yaml["names"].values())  # names list
+          for obj in root.iter("object"):
+              cls = obj.find("name").text
+              if cls in names and int(obj.find("difficult").text) != 1:
+                  xmlbox = obj.find("bndbox")
+                  bb = convert_box((w, h), [float(xmlbox.find(x).text) for x in ("xmin", "xmax", "ymin", "ymax")])
+                  cls_id = names.index(cls)  # class id
+                  out_file.write(" ".join(str(a) for a in (cls_id, *bb)) + "\n")
 
 
   # Download

ultralytics/cfg/datasets/african-wildlife.yaml

@@ -1,6 +1,6 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-# African-wildlife dataset by Ultralytics
+# African Wildlife dataset by Ultralytics
 # Documentation: https://docs.ultralytics.com/datasets/detect/african-wildlife/
 # Example usage: yolo train data=african-wildlife.yaml
 # parent

ultralytics/cfg/datasets/coco128-seg.yaml

@@ -2,7 +2,7 @@
 
 # COCO128-seg dataset https://www.kaggle.com/datasets/ultralytics/coco128 (first 128 images from COCO train2017) by Ultralytics
 # Documentation: https://docs.ultralytics.com/datasets/segment/coco/
-# Example usage: yolo train data=coco128.yaml
+# Example usage: yolo train data=coco128-seg.yaml
 # parent
 # ├── ultralytics
 # └── datasets

ultralytics/cfg/datasets/dota8-multispectral.yaml

@@ -6,7 +6,7 @@
 # parent
 # ├── ultralytics
 # └── datasets
-#     └── dota8-multispectral ← downloads here (37.3MB)
+#     └── dota8-multispectral ← downloads here (37.3 MB)
 
 # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]
 path: dota8-multispectral # dataset root dir

ultralytics/cfg/datasets/dota8.yaml

@@ -1,12 +1,12 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-# DOTA8 dataset 8 images from split DOTAv1 dataset by Ultralytics
+# DOTA8 dataset (8 images from the DOTAv1 split) by Ultralytics
 # Documentation: https://docs.ultralytics.com/datasets/obb/dota8/
 # Example usage: yolo train model=yolov8n-obb.pt data=dota8.yaml
 # parent
 # ├── ultralytics
 # └── datasets
-#     └── dota8 ← downloads here (1MB)
+#     └── dota8 ← downloads here (1 MB)
 
 # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]
 path: dota8 # dataset root dir

ultralytics/cfg/datasets/kitti.yaml

@@ -1,6 +1,6 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-# Kitti dataset by Karlsruhe Institute of Technology and Toyota Technological Institute at Chicago
+# KITTI dataset by Karlsruhe Institute of Technology and Toyota Technological Institute at Chicago
 # Documentation: https://docs.ultralytics.com/datasets/detect/kitti/
 # Example usage: yolo train data=kitti.yaml
 # parent

ultralytics/cfg/datasets/xView.yaml

@@ -1,7 +1,7 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-# DIUx xView 2018 Challenge https://challenge.xviewdataset.org by U.S. National Geospatial-Intelligence Agency (NGA)
-# --------  DOWNLOAD DATA MANUALLY and jar xf val_images.zip to 'datasets/xView' before running train command!  --------
+# DIUx xView 2018 Challenge dataset https://challenge.xviewdataset.org by U.S. National Geospatial-Intelligence Agency (NGA)
+# --------  Download and extract data manually to `datasets/xView` before running the train command.  --------
 # Documentation: https://docs.ultralytics.com/datasets/detect/xview/
 # Example usage: yolo train data=xView.yaml
 # parent
@@ -12,7 +12,7 @@
 # Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]
 path: xView # dataset root dir
 train: images/autosplit_train.txt # train images (relative to 'path') 90% of 847 train images
-val: images/autosplit_val.txt # train images (relative to 'path') 10% of 847 train images
+val: images/autosplit_val.txt # val images (relative to 'path') 10% of 847 train images
 
 # Classes
 names:
@@ -80,8 +80,8 @@ names:
 # Download script/URL (optional) ---------------------------------------------------------------------------------------
 download: |
   import json
-  import os
   from pathlib import Path
+  import shutil
 
   import numpy as np
   from PIL import Image
@@ -92,15 +92,15 @@ download: |
 
 
   def convert_labels(fname=Path("xView/xView_train.geojson")):
-      """Converts xView geoJSON labels to YOLO format, mapping classes to indices 0-59 and saving as text files."""
+      """Convert xView GeoJSON labels to YOLO format (classes 0-59) and save them as text files."""
       path = fname.parent
       with open(fname, encoding="utf-8") as f:
           print(f"Loading {fname}...")
           data = json.load(f)
 
       # Make dirs
-      labels = Path(path / "labels" / "train")
-      os.system(f"rm -rf {labels}")
+      labels = path / "labels" / "train"
+      shutil.rmtree(labels, ignore_errors=True)
       labels.mkdir(parents=True, exist_ok=True)
 
       # xView classes 11-94 to 0-59
@@ -113,24 +113,24 @@ download: |
       for feature in TQDM(data["features"], desc=f"Converting {fname}"):
           p = feature["properties"]
           if p["bounds_imcoords"]:
-              id = p["image_id"]
-              file = path / "train_images" / id
-              if file.exists():  # 1395.tif missing
+              image_id = p["image_id"]
+              image_file = path / "train_images" / image_id
+              if image_file.exists():  # 1395.tif missing
                   try:
                       box = np.array([int(num) for num in p["bounds_imcoords"].split(",")])
                       assert box.shape[0] == 4, f"incorrect box shape {box.shape[0]}"
                       cls = p["type_id"]
-                      cls = xview_class2index[int(cls)]  # xView class to 0-60
+                      cls = xview_class2index[int(cls)]  # xView class to 0-59
                       assert 59 >= cls >= 0, f"incorrect class index {cls}"
 
                       # Write YOLO label
-                      if id not in shapes:
-                          shapes[id] = Image.open(file).size
-                      box = xyxy2xywhn(box[None].astype(np.float), w=shapes[id][0], h=shapes[id][1], clip=True)
-                      with open((labels / id).with_suffix(".txt"), "a", encoding="utf-8") as f:
+                      if image_id not in shapes:
+                          shapes[image_id] = Image.open(image_file).size
+                      box = xyxy2xywhn(box[None].astype(float), w=shapes[image_id][0], h=shapes[image_id][1], clip=True)
+                      with open((labels / image_id).with_suffix(".txt"), "a", encoding="utf-8") as f:
                           f.write(f"{cls} {' '.join(f'{x:.6f}' for x in box[0])}\n")  # write label.txt
                   except Exception as e:
-                      print(f"WARNING: skipping one label for {file}: {e}")
+                      print(f"WARNING: skipping one label for {image_file}: {e}")
 
 
   # Download manually from https://challenge.xviewdataset.org

ultralytics/cfg/models/11/yolo11-pose.yaml

@@ -7,7 +7,7 @@
 # Parameters
 nc: 80 # number of classes
 kpt_shape: [17, 3] # number of keypoints, number of dims (2 for x,y or 3 for x,y,visible)
-scales: # model compound scaling constants, i.e. 'model=yolo11n-pose.yaml' will call yolo11.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolo11n-pose.yaml' will call yolo11-pose.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.50, 0.25, 1024] # summary: 196 layers, 2908507 parameters, 2908491 gradients, 7.7 GFLOPs
   s: [0.50, 0.50, 1024] # summary: 196 layers, 9948811 parameters, 9948795 gradients, 23.5 GFLOPs

ultralytics/cfg/models/11/yoloe-11-seg.yaml

@@ -1,10 +1,10 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-# YOLO11-seg instance segmentation model. For Usage examples see https://docs.ultralytics.com/tasks/segment
+# Ultralytics YOLOE-11-seg instance segmentation model. For usage examples, see https://docs.ultralytics.com/tasks/segment
 
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolo11n-seg.yaml' will call yolo11-seg.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yoloe-11n-seg.yaml' will call yoloe-11-seg.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.50, 0.25, 1024] # summary: 355 layers, 2876848 parameters, 2876832 gradients, 10.5 GFLOPs
   s: [0.50, 0.50, 1024] # summary: 355 layers, 10113248 parameters, 10113232 gradients, 35.8 GFLOPs

ultralytics/cfg/models/11/yoloe-11.yaml

@@ -1,10 +1,10 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-# YOLO11 object detection model with P3-P5 outputs. For Usage examples see https://docs.ultralytics.com/tasks/detect
+# Ultralytics YOLOE-11 object detection model with P3/8 - P5/32 outputs. For usage examples, see https://docs.ultralytics.com/tasks/detect
 
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolo11n.yaml' will call yolo11.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yoloe-11n.yaml' will call yoloe-11.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.50, 0.25, 1024] # summary: 319 layers, 2624080 parameters, 2624064 gradients, 6.6 GFLOPs
   s: [0.50, 0.50, 1024] # summary: 319 layers, 9458752 parameters, 9458736 gradients, 21.7 GFLOPs

ultralytics/cfg/models/v8/yoloe-v8-seg.yaml

@@ -1,14 +1,17 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+# Ultralytics YOLOE-v8-seg instance segmentation model with P3/8 - P5/32 outputs
+# Task docs: https://docs.ultralytics.com/tasks/segment
+
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call yolov8.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yoloe-v8n-seg.yaml' will call yoloe-v8-seg.yaml with scale 'n'
   # [depth, width, max_channels]
-  n: [0.33, 0.25, 1024] # YOLOv8n-world summary: 161 layers, 4204111 parameters, 4204095 gradients, 39.6 GFLOPs
-  s: [0.33, 0.50, 1024] # YOLOv8s-world summary: 161 layers, 13383496 parameters, 13383480 gradients, 71.5 GFLOPs
-  m: [0.67, 0.75, 768] # YOLOv8m-world summary: 201 layers, 29065310 parameters, 29065294 gradients, 131.4 GFLOPs
-  l: [1.00, 1.00, 512] # YOLOv8l-world summary: 241 layers, 47553970 parameters, 47553954 gradients, 225.6 GFLOPs
-  x: [1.00, 1.25, 512] # YOLOv8x-world summary: 241 layers, 73690217 parameters, 73690201 gradients, 330.8 GFLOPs
+  n: [0.33, 0.25, 1024] # YOLOE-v8n-seg summary: 161 layers, 4204111 parameters, 4204095 gradients, 39.6 GFLOPs
+  s: [0.33, 0.50, 1024] # YOLOE-v8s-seg summary: 161 layers, 13383496 parameters, 13383480 gradients, 71.5 GFLOPs
+  m: [0.67, 0.75, 768] # YOLOE-v8m-seg summary: 201 layers, 29065310 parameters, 29065294 gradients, 131.4 GFLOPs
+  l: [1.00, 1.00, 512] # YOLOE-v8l-seg summary: 241 layers, 47553970 parameters, 47553954 gradients, 225.6 GFLOPs
+  x: [1.00, 1.25, 512] # YOLOE-v8x-seg summary: 241 layers, 73690217 parameters, 73690201 gradients, 330.8 GFLOPs
 
 # YOLOv8.0n backbone
 backbone:

ultralytics/cfg/models/v8/yoloe-v8.yaml

@@ -1,14 +1,17 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+# Ultralytics YOLOE-v8 object detection model with P3/8 - P5/32 outputs
+# Task docs: https://docs.ultralytics.com/tasks/detect
+
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call yolov8.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yoloe-v8n.yaml' will call yoloe-v8.yaml with scale 'n'
   # [depth, width, max_channels]
-  n: [0.33, 0.25, 1024] # YOLOv8n-worldv2 summary: 148 layers, 3695183 parameters, 3695167 gradients, 19.5 GFLOPS
-  s: [0.33, 0.50, 1024] # YOLOv8s-worldv2 summary: 148 layers, 12759880 parameters, 12759864 gradients, 51.0 GFLOPS
-  m: [0.67, 0.75, 768] # YOLOv8m-worldv2 summary: 188 layers, 28376158 parameters, 28376142 gradients, 110.5 GFLOPS
-  l: [1.00, 1.00, 512] # YOLOv8l-worldv2 summary: 228 layers, 46832050 parameters, 46832034 gradients, 204.5 GFLOPS
-  x: [1.00, 1.25, 512] # YOLOv8x-worldv2 summary: 228 layers, 72886377 parameters, 72886361 gradients, 309.3 GFLOPS
+  n: [0.33, 0.25, 1024] # YOLOE-v8n summary: 148 layers, 3695183 parameters, 3695167 gradients, 19.5 GFLOPs
+  s: [0.33, 0.50, 1024] # YOLOE-v8s summary: 148 layers, 12759880 parameters, 12759864 gradients, 51.0 GFLOPs
+  m: [0.67, 0.75, 768] # YOLOE-v8m summary: 188 layers, 28376158 parameters, 28376142 gradients, 110.5 GFLOPs
+  l: [1.00, 1.00, 512] # YOLOE-v8l summary: 228 layers, 46832050 parameters, 46832034 gradients, 204.5 GFLOPs
+  x: [1.00, 1.25, 512] # YOLOE-v8x summary: 228 layers, 72886377 parameters, 72886361 gradients, 309.3 GFLOPs
 
 # YOLOv8.0n backbone
 backbone:

ultralytics/cfg/models/v8/yolov8-cls-resnet101.yaml

@@ -6,7 +6,7 @@
 
 # Parameters
 nc: 1000 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n-cls.yaml' will call yolov8-cls.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-cls-resnet101.yaml' will call yolov8-cls-resnet101.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.33, 0.25, 1024]
   s: [0.33, 0.50, 1024]

ultralytics/cfg/models/v8/yolov8-cls-resnet50.yaml

@@ -6,7 +6,7 @@
 
 # Parameters
 nc: 1000 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n-cls.yaml' will call yolov8-cls.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-cls-resnet50.yaml' will call yolov8-cls-resnet50.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.33, 0.25, 1024]
   s: [0.33, 0.50, 1024]

ultralytics/cfg/models/v8/yolov8-ghost-p2.yaml

@@ -1,13 +1,13 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-# Ultralytics YOLOv8 object detection model with P2/4 - P5/32 outputs
+# Ultralytics YOLOv8-ghost object detection model with P2/4 - P5/32 outputs
 # Model docs: https://docs.ultralytics.com/models/yolov8
 # Task docs: https://docs.ultralytics.com/tasks/detect
 # Employs Ghost convolutions and modules proposed in Huawei's GhostNet in https://arxiv.org/abs/1911.11907v2
 
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call yolov8.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-ghost-p2.yaml' will call yolov8-ghost-p2.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.33, 0.25, 1024] # YOLOv8n-ghost-p2 summary: 290 layers, 2033944 parameters, 2033928 gradients, 13.8 GFLOPs
   s: [0.33, 0.50, 1024] # YOLOv8s-ghost-p2 summary: 290 layers, 5562080 parameters, 5562064 gradients, 25.1 GFLOPs

ultralytics/cfg/models/v8/yolov8-ghost-p6.yaml

@@ -1,13 +1,13 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-# Ultralytics YOLOv8 object detection model with P3/8 - P6/64 outputs
+# Ultralytics YOLOv8-ghost object detection model with P3/8 - P6/64 outputs
 # Model docs: https://docs.ultralytics.com/models/yolov8
 # Task docs: https://docs.ultralytics.com/tasks/detect
 # Employs Ghost convolutions and modules proposed in Huawei's GhostNet in https://arxiv.org/abs/1911.11907v2
 
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n-p6.yaml' will call yolov8-p6.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-ghost-p6.yaml' will call yolov8-ghost-p6.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.33, 0.25, 1024] # YOLOv8n-ghost-p6 summary: 312 layers, 2901100 parameters, 2901084 gradients, 5.8 GFLOPs
   s: [0.33, 0.50, 1024] # YOLOv8s-ghost-p6 summary: 312 layers, 9520008 parameters, 9519992 gradients, 16.4 GFLOPs

ultralytics/cfg/models/v8/yolov8-ghost.yaml

@@ -1,13 +1,13 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-# Ultralytics YOLOv8 object detection model with P3/8 - P5/32 outputs
+# Ultralytics YOLOv8-ghost object detection model with P3/8 - P5/32 outputs
 # Model docs: https://docs.ultralytics.com/models/yolov8
 # Task docs: https://docs.ultralytics.com/tasks/detect
 # Employs Ghost convolutions and modules proposed in Huawei's GhostNet in https://arxiv.org/abs/1911.11907v2
 
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call yolov8.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-ghost.yaml' will call yolov8-ghost.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.33, 0.25, 1024] # YOLOv8n-ghost summary: 237 layers, 1865316 parameters, 1865300 gradients, 5.8 GFLOPs
   s: [0.33, 0.50, 1024] # YOLOv8s-ghost summary: 237 layers, 5960072 parameters, 5960056 gradients, 16.4 GFLOPs

ultralytics/cfg/models/v8/yolov8-obb.yaml

@@ -6,7 +6,7 @@
 
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call yolov8.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-obb.yaml' will call yolov8-obb.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.33, 0.25, 1024] # YOLOv8n-obb summary: 144 layers, 3228867 parameters, 3228851 gradients, 9.1 GFLOPs
   s: [0.33, 0.50, 1024] # YOLOv8s-obb summary: 144 layers, 11452739 parameters, 11452723 gradients, 29.8 GFLOPs

ultralytics/cfg/models/v8/yolov8-p2.yaml

@@ -6,7 +6,7 @@
 
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call yolov8.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-p2.yaml' will call yolov8-p2.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.33, 0.25, 1024]
   s: [0.33, 0.50, 1024]

ultralytics/cfg/models/v8/yolov8-pose-p6.yaml

@@ -7,7 +7,7 @@
 # Parameters
 nc: 1 # number of classes
 kpt_shape: [17, 3] # number of keypoints, number of dims (2 for x,y or 3 for x,y,visible)
-scales: # model compound scaling constants, i.e. 'model=yolov8n-p6.yaml' will call yolov8-p6.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-pose-p6.yaml' will call yolov8-pose-p6.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.33, 0.25, 1024]
   s: [0.33, 0.50, 1024]

ultralytics/cfg/models/v8/yolov8-rtdetr.yaml

@@ -6,7 +6,7 @@
 
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call yolov8.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-rtdetr.yaml' will call yolov8-rtdetr.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.33, 0.25, 1024] # YOLOv8n-rtdetr summary: 235 layers, 9643868 parameters, 9643868 gradients, 17.1 GFLOPs
   s: [0.33, 0.50, 1024] # YOLOv8s-rtdetr summary: 235 layers, 16518572 parameters, 16518572 gradients, 32.8 GFLOPs

ultralytics/cfg/models/v8/yolov8-world.yaml

@@ -6,7 +6,7 @@
 
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call yolov8.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-world.yaml' will call yolov8-world.yaml with scale 'n'
   # [depth, width, max_channels]
   n: [0.33, 0.25, 1024] # YOLOv8n-world summary: 161 layers, 4204111 parameters, 4204095 gradients, 39.6 GFLOPs
   s: [0.33, 0.50, 1024] # YOLOv8s-world summary: 161 layers, 13383496 parameters, 13383480 gradients, 71.5 GFLOPs

ultralytics/cfg/models/v8/yolov8-worldv2.yaml

@@ -6,13 +6,13 @@
 
 # Parameters
 nc: 80 # number of classes
-scales: # model compound scaling constants, i.e. 'model=yolov8n.yaml' will call yolov8.yaml with scale 'n'
+scales: # model compound scaling constants, i.e. 'model=yolov8n-worldv2.yaml' will call yolov8-worldv2.yaml with scale 'n'
   # [depth, width, max_channels]
-  n: [0.33, 0.25, 1024] # YOLOv8n-worldv2 summary: 148 layers, 3695183 parameters, 3695167 gradients, 19.5 GFLOPS
-  s: [0.33, 0.50, 1024] # YOLOv8s-worldv2 summary: 148 layers, 12759880 parameters, 12759864 gradients, 51.0 GFLOPS
-  m: [0.67, 0.75, 768] # YOLOv8m-worldv2 summary: 188 layers, 28376158 parameters, 28376142 gradients, 110.5 GFLOPS
-  l: [1.00, 1.00, 512] # YOLOv8l-worldv2 summary: 228 layers, 46832050 parameters, 46832034 gradients, 204.5 GFLOPS
-  x: [1.00, 1.25, 512] # YOLOv8x-worldv2 summary: 228 layers, 72886377 parameters, 72886361 gradients, 309.3 GFLOPS
+  n: [0.33, 0.25, 1024] # YOLOv8n-worldv2 summary: 148 layers, 3695183 parameters, 3695167 gradients, 19.5 GFLOPs
+  s: [0.33, 0.50, 1024] # YOLOv8s-worldv2 summary: 148 layers, 12759880 parameters, 12759864 gradients, 51.0 GFLOPs
+  m: [0.67, 0.75, 768] # YOLOv8m-worldv2 summary: 188 layers, 28376158 parameters, 28376142 gradients, 110.5 GFLOPs
+  l: [1.00, 1.00, 512] # YOLOv8l-worldv2 summary: 228 layers, 46832050 parameters, 46832034 gradients, 204.5 GFLOPs
+  x: [1.00, 1.25, 512] # YOLOv8x-worldv2 summary: 228 layers, 72886377 parameters, 72886361 gradients, 309.3 GFLOPs
 
 # YOLOv8.0n backbone
 backbone:

ultralytics/data/augment.py

@@ -1697,7 +1697,7 @@ class CopyPaste(BaseMixTransform):
     """
 
     def __init__(self, dataset=None, pre_transform=None, p: float = 0.5, mode: str = "flip") -> None:
-        """Initialize CopyPaste object with dataset, pre_transform, and probability of applying MixUp."""
+        """Initialize CopyPaste object with dataset, pre_transform, and probability of applying CopyPaste."""
         super().__init__(dataset=dataset, pre_transform=pre_transform, p=p)
         assert mode in {"flip", "mixup"}, f"Expected `mode` to be `flip` or `mixup`, but got {mode}."
         self.mode = mode

ultralytics/data/base.py

@@ -33,7 +33,8 @@ class BaseDataset(Dataset):
         single_cls (bool): Whether to treat all objects as a single class.
         prefix (str): Prefix to print in log messages.
         fraction (float): Fraction of dataset to utilize.
-        channels (int): Number of channels in the images (1 for grayscale, 3 for RGB).
+        channels (int): Number of channels in the images (1 for grayscale, 3 for color). Color images loaded with OpenCV
+            are in BGR channel order.
         cv2_flag (int): OpenCV flag for reading images.
         im_files (list[str]): List of image file paths.
         labels (list[dict]): List of label data dictionaries.
@@ -101,7 +102,8 @@ class BaseDataset(Dataset):
             single_cls (bool): If True, single class training is used.
             classes (list[int], optional): List of included classes.
             fraction (float): Fraction of dataset to utilize.
-            channels (int): Number of channels in the images (1 for grayscale, 3 for RGB).
+            channels (int): Number of channels in the images (1 for grayscale, 3 for color). Color images loaded with
+                OpenCV are in BGR channel order.
         """
         super().__init__()
         self.img_path = img_path

ultralytics/data/build.py

@@ -35,7 +35,7 @@ from ultralytics.utils.torch_utils import TORCH_2_0
 
 
 class InfiniteDataLoader(dataloader.DataLoader):
-    """Dataloader that reuses workers for infinite iteration.
+    """DataLoader that reuses workers for infinite iteration.
 
     This dataloader extends the PyTorch DataLoader to provide infinite recycling of workers, which improves efficiency
     for training loops that need to iterate through the dataset multiple times without recreating workers.
@@ -51,7 +51,7 @@ class InfiniteDataLoader(dataloader.DataLoader):
         reset: Reset the iterator, useful when modifying dataset settings during training.
 
     Examples:
-        Create an infinite dataloader for training
+        Create an infinite DataLoader for training
         >>> dataset = YOLODataset(...)
         >>> dataloader = InfiniteDataLoader(dataset, batch_size=16, shuffle=True)
         >>> for batch in dataloader:  # Infinite iteration
@@ -76,7 +76,7 @@ class InfiniteDataLoader(dataloader.DataLoader):
             yield next(self.iterator)
 
     def __del__(self):
-        """Ensure that workers are properly terminated when the dataloader is deleted."""
+        """Ensure that workers are properly terminated when the DataLoader is deleted."""
         try:
             if not hasattr(self.iterator, "_workers"):
                 return
@@ -417,7 +417,7 @@ def load_inference_source(
     source, stream, screenshot, from_img, in_memory, tensor = check_source(source)
     source_type = source.source_type if in_memory else SourceTypes(stream, screenshot, from_img, tensor)
 
-    # Dataloader
+    # DataLoader
     if tensor:
         dataset = LoadTensor(source)
     elif in_memory:

ultralytics/data/loaders.py

@@ -69,7 +69,7 @@ class LoadStreams:
         shape (list[tuple[int, int, int]]): List of shapes for each stream.
         caps (list[cv2.VideoCapture]): List of cv2.VideoCapture objects for each stream.
         bs (int): Batch size for processing.
-        cv2_flag (int): OpenCV flag for image reading (grayscale or RGB).
+        cv2_flag (int): OpenCV flag for image reading (grayscale or color/BGR).
 
     Methods:
         update: Read stream frames in daemon thread.
@@ -98,14 +98,14 @@ class LoadStreams:
             sources (str): Path to streams file or single stream URL.
             vid_stride (int): Video frame-rate stride.
             buffer (bool): Whether to buffer input streams.
-            channels (int): Number of image channels (1 for grayscale, 3 for RGB).
+            channels (int): Number of image channels (1 for grayscale, 3 for color).
         """
         torch.backends.cudnn.benchmark = True  # faster for fixed-size inference
         self.buffer = buffer  # buffer input streams
         self.running = True  # running flag for Thread
         self.mode = "stream"
         self.vid_stride = vid_stride  # video frame-rate stride
-        self.cv2_flag = cv2.IMREAD_GRAYSCALE if channels == 1 else cv2.IMREAD_COLOR  # grayscale or RGB
+        self.cv2_flag = cv2.IMREAD_GRAYSCALE if channels == 1 else cv2.IMREAD_COLOR  # grayscale or color (BGR)
 
         sources = Path(sources).read_text().rsplit() if os.path.isfile(sources) else [sources]
         n = len(sources)
@@ -242,7 +242,7 @@ class LoadScreenshots:
         bs (int): Batch size, set to 1.
         fps (int): Frames per second, set to 30.
         monitor (dict[str, int]): Monitor configuration details.
-        cv2_flag (int): OpenCV flag for image reading (grayscale or RGB).
+        cv2_flag (int): OpenCV flag for image reading (grayscale or color/BGR).
 
     Methods:
         __iter__: Returns an iterator object.
@@ -259,7 +259,7 @@ class LoadScreenshots:
 
         Args:
             source (str): Screen capture source string in format "screen_num left top width height".
-            channels (int): Number of image channels (1 for grayscale, 3 for RGB).
+            channels (int): Number of image channels (1 for grayscale, 3 for color).
         """
         check_requirements("mss")
         import mss
@@ -277,7 +277,7 @@ class LoadScreenshots:
         self.sct = mss.mss()
         self.bs = 1
         self.fps = 30
-        self.cv2_flag = cv2.IMREAD_GRAYSCALE if channels == 1 else cv2.IMREAD_COLOR  # grayscale or RGB
+        self.cv2_flag = cv2.IMREAD_GRAYSCALE if channels == 1 else cv2.IMREAD_COLOR  # grayscale or color (BGR)
 
         # Parse monitor shape
         monitor = self.sct.monitors[self.screen]
@@ -319,7 +319,7 @@ class LoadImagesAndVideos:
         frames (int): Total number of frames in the video.
         count (int): Counter for iteration, initialized at 0 during __iter__().
         ni (int): Number of images.
-        cv2_flag (int): OpenCV flag for image reading (grayscale or RGB).
+        cv2_flag (int): OpenCV flag for image reading (grayscale or color/BGR).
 
     Methods:
         __init__: Initialize the LoadImagesAndVideos object.
@@ -347,7 +347,7 @@ class LoadImagesAndVideos:
             path (str | Path | list): Path to images/videos, directory, or list of paths.
             batch (int): Batch size for processing.
             vid_stride (int): Video frame-rate stride.
-            channels (int): Number of image channels (1 for grayscale, 3 for RGB).
+            channels (int): Number of image channels (1 for grayscale, 3 for color).
         """
         parent = None
         if isinstance(path, str) and Path(path).suffix in {".txt", ".csv"}:  # txt/csv file with source paths
@@ -385,7 +385,7 @@ class LoadImagesAndVideos:
         self.mode = "video" if ni == 0 else "image"  # default to video if no images
         self.vid_stride = vid_stride  # video frame-rate stride
         self.bs = batch
-        self.cv2_flag = cv2.IMREAD_GRAYSCALE if channels == 1 else cv2.IMREAD_COLOR  # grayscale or RGB
+        self.cv2_flag = cv2.IMREAD_GRAYSCALE if channels == 1 else cv2.IMREAD_COLOR  # grayscale or color (BGR)
         if any(videos):
             self._new_video(videos[0])  # new video
         else:
@@ -513,7 +513,7 @@ class LoadPilAndNumpy:
 
         Args:
             im0 (PIL.Image.Image | np.ndarray | list): Single image or list of images in PIL or numpy format.
-            channels (int): Number of image channels (1 for grayscale, 3 for RGB).
+            channels (int): Number of image channels (1 for grayscale, 3 for color).
         """
         if not isinstance(im0, list):
             im0 = [im0]
@@ -526,11 +526,16 @@ class LoadPilAndNumpy:
 
     @staticmethod
     def _single_check(im: Image.Image | np.ndarray, flag: str = "RGB") -> np.ndarray:
-        """Validate and format an image to numpy array, ensuring RGB order and contiguous memory."""
+        """Validate and format an image to a NumPy array.
+
+        Notes:
+            - PIL inputs are converted to NumPy and returned in OpenCV-compatible BGR order for color images.
+            - NumPy inputs are returned as-is (no channel-order conversion is applied).
+        """
         assert isinstance(im, (Image.Image, np.ndarray)), f"Expected PIL/np.ndarray image type, but got {type(im)}"
         if isinstance(im, Image.Image):
             im = np.asarray(im.convert(flag))
-            # adding new axis if it's grayscale, and converting to BGR if it's RGB
+            # Add a new axis if grayscale; convert RGB -> BGR for OpenCV compatibility.
             im = im[..., None] if flag == "L" else im[..., ::-1]
             im = np.ascontiguousarray(im)  # contiguous
         elif im.ndim == 2:  # grayscale in numpy form

ultralytics/data/utils.py

@@ -258,12 +258,12 @@ def visualize_image_annotations(image_path: str, txt_path: str, label_map: dict[
     luminance.
 
     Args:
-        image_path (str): The path to the image file to annotate, and it can be in formats supported by PIL.
-        txt_path (str): The path to the annotation file in YOLO format, that should contain one line per object.
+        image_path (str): Path to the image file to annotate. The file must be readable by PIL.
+        txt_path (str): Path to the annotation file in YOLO format, which should contain one line per object.
         label_map (dict[int, str]): A dictionary that maps class IDs (integers) to class labels (strings).
 
     Examples:
-        >>> label_map = {0: "cat", 1: "dog", 2: "bird"}  # It should include all annotated classes details
+        >>> label_map = {0: "cat", 1: "dog", 2: "bird"}  # Should include all annotated classes
         >>> visualize_image_annotations("path/to/image.jpg", "path/to/annotations.txt", label_map)
     """
     import matplotlib.pyplot as plt
@@ -283,7 +283,7 @@ def visualize_image_annotations(image_path: str, txt_path: str, label_map: dict[
             annotations.append((x, y, w, h, int(class_id)))
     _, ax = plt.subplots(1)  # Plot the image and annotations
     for x, y, w, h, label in annotations:
-        color = tuple(c / 255 for c in colors(label, True))  # Get and normalize the RGB color
+        color = tuple(c / 255 for c in colors(label, False))  # Get and normalize an RGB color for Matplotlib
         rect = plt.Rectangle((x, y), w, h, linewidth=2, edgecolor=color, facecolor="none")  # Create a rectangle
         ax.add_patch(rect)
         luminance = 0.2126 * color[0] + 0.7152 * color[1] + 0.0722 * color[2]  # Formula for luminance