dgenerate-ultralytics-headless 8.3.159__tar.gz → 8.3.160__tar.gz

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dgenerate-ultralytics-headless
-Version: 8.3.159
+Version: 8.3.160
 Summary: Automatically built Ultralytics package with python-opencv-headless dependency instead of python-opencv
 Author-email: Glenn Jocher <glenn.jocher@ultralytics.com>, Jing Qiu <jing.qiu@ultralytics.com>
 Maintainer-email: Ultralytics <hello@ultralytics.com>

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/dgenerate_ultralytics_headless.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dgenerate-ultralytics-headless
-Version: 8.3.159
+Version: 8.3.160
 Summary: Automatically built Ultralytics package with python-opencv-headless dependency instead of python-opencv
 Author-email: Glenn Jocher <glenn.jocher@ultralytics.com>, Jing Qiu <jing.qiu@ultralytics.com>
 Maintainer-email: Ultralytics <hello@ultralytics.com>

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/__init__.py

@@ -1,6 +1,6 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-__version__ = "8.3.159"
+__version__ = "8.3.160"
 
 import os
 

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/cfg/__init__.py

@@ -954,8 +954,6 @@ def entrypoint(debug: str = "") -> None:
         from ultralytics import YOLO
 
         model = YOLO(model, task=task)
-    if isinstance(overrides.get("pretrained"), str):
-        model.load(overrides["pretrained"])
 
     # Task Update
     if task != model.task:

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/data/augment.py

@@ -251,8 +251,7 @@ class Compose:
             >>> multiple_transforms = compose[0:2]  # Returns a Compose object with RandomFlip and RandomPerspective
         """
         assert isinstance(index, (int, list)), f"The indices should be either list or int type but got {type(index)}"
-        index = [index] if isinstance(index, int) else index
-        return Compose([self.transforms[i] for i in index])
+        return Compose([self.transforms[i] for i in index]) if isinstance(index, list) else self.transforms[index]
 
     def __setitem__(self, index: Union[list, int], value: Union[list, int]) -> None:
         """
@@ -1560,14 +1559,15 @@ class RandomFlip:
         h = 1 if instances.normalized else h
         w = 1 if instances.normalized else w
 
-        # Flip up-down
+        # WARNING: two separate if and calls to random.random() intentional for reproducibility with older versions
         if self.direction == "vertical" and random.random() < self.p:
             img = np.flipud(img)
             instances.flipud(h)
+            if self.flip_idx is not None and instances.keypoints is not None:
+                instances.keypoints = np.ascontiguousarray(instances.keypoints[:, self.flip_idx, :])
         if self.direction == "horizontal" and random.random() < self.p:
             img = np.fliplr(img)
             instances.fliplr(w)
-            # For keypoints
             if self.flip_idx is not None and instances.keypoints is not None:
                 instances.keypoints = np.ascontiguousarray(instances.keypoints[:, self.flip_idx, :])
         labels["img"] = np.ascontiguousarray(img)
@@ -2533,9 +2533,9 @@ def v8_transforms(dataset, imgsz, hyp, stretch=False):
     flip_idx = dataset.data.get("flip_idx", [])  # for keypoints augmentation
     if dataset.use_keypoints:
         kpt_shape = dataset.data.get("kpt_shape", None)
-        if len(flip_idx) == 0 and hyp.fliplr > 0.0:
-            hyp.fliplr = 0.0
-            LOGGER.warning("No 'flip_idx' array defined in data.yaml, setting augmentation 'fliplr=0.0'")
+        if len(flip_idx) == 0 and (hyp.fliplr > 0.0 or hyp.flipud > 0.0):
+            hyp.fliplr = hyp.flipud = 0.0  # both fliplr and flipud require flip_idx
+            LOGGER.warning("No 'flip_idx' array defined in data.yaml, disabling 'fliplr' and 'flipud' augmentations.")
         elif flip_idx and (len(flip_idx) != kpt_shape[0]):
             raise ValueError(f"data.yaml flip_idx={flip_idx} length must be equal to kpt_shape[0]={kpt_shape[0]}")
 
@@ -2546,7 +2546,7 @@ def v8_transforms(dataset, imgsz, hyp, stretch=False):
             CutMix(dataset, pre_transform=pre_transform, p=hyp.cutmix),
             Albumentations(p=1.0),
             RandomHSV(hgain=hyp.hsv_h, sgain=hyp.hsv_s, vgain=hyp.hsv_v),
-            RandomFlip(direction="vertical", p=hyp.flipud),
+            RandomFlip(direction="vertical", p=hyp.flipud, flip_idx=flip_idx),
             RandomFlip(direction="horizontal", p=hyp.fliplr, flip_idx=flip_idx),
         ]
     )  # transforms

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/engine/exporter.py

@@ -1495,7 +1495,7 @@ class NMSModel(torch.nn.Module):
         scores, classes = scores.max(dim=-1)
         self.args.max_det = min(pred.shape[1], self.args.max_det)  # in case num_anchors < max_det
         # (N, max_det, 4 coords + 1 class score + 1 class label + extra_shape).
-        out = torch.zeros(bs, self.args.max_det, boxes.shape[-1] + 2 + extra_shape, **kwargs)
+        out = torch.zeros(pred.shape[0], self.args.max_det, boxes.shape[-1] + 2 + extra_shape, **kwargs)
         for i in range(bs):
             box, cls, score, extra = boxes[i], classes[i], scores[i], extras[i]
             mask = score > self.args.conf

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/engine/model.py

@@ -777,6 +777,8 @@ class Model(torch.nn.Module):
 
         checks.check_pip_update_available()
 
+        if isinstance(kwargs.get("pretrained", None), (str, Path)):
+            self.load(kwargs["pretrained"])  # load pretrained weights if provided
         overrides = YAML.load(checks.check_yaml(kwargs["cfg"])) if kwargs.get("cfg") else self.overrides
         custom = {
             # NOTE: handle the case when 'cfg' includes 'data'.

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/engine/results.py

@@ -16,7 +16,6 @@ import torch
 from ultralytics.data.augment import LetterBox
 from ultralytics.utils import LOGGER, DataExportMixin, SimpleClass, ops
 from ultralytics.utils.plotting import Annotator, colors, save_one_box
-from ultralytics.utils.torch_utils import smart_inference_mode
 
 
 class BaseTensor(SimpleClass):
@@ -1204,7 +1203,6 @@ class Keypoints(BaseTensor):
         >>> keypoints_cpu = keypoints.cpu()  # Move keypoints to CPU
     """
 
-    @smart_inference_mode()  # avoid keypoints < conf in-place error
     def __init__(self, keypoints: Union[torch.Tensor, np.ndarray], orig_shape: Tuple[int, int]) -> None:
         """
         Initialize the Keypoints object with detection keypoints and original image dimensions.
@@ -1225,9 +1223,6 @@ class Keypoints(BaseTensor):
         """
         if keypoints.ndim == 2:
             keypoints = keypoints[None, :]
-        if keypoints.shape[2] == 3:  # x, y, conf
-            mask = keypoints[..., 2] < 0.5  # points with conf < 0.5 (not visible)
-            keypoints[..., :2][mask] = 0
         super().__init__(keypoints, orig_shape)
         self.has_visible = self.data.shape[-1] == 3
 

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/models/yolo/model.py

@@ -406,18 +406,18 @@ class YOLOE(Model):
                 f"Expected equal number of bounding boxes and classes, but got {len(visual_prompts['bboxes'])} and "
                 f"{len(visual_prompts['cls'])} respectively"
             )
-        self.predictor = (predictor or self._smart_load("predictor"))(
-            overrides={
-                "task": self.model.task,
-                "mode": "predict",
-                "save": False,
-                "verbose": refer_image is None,
-                "batch": 1,
-            },
-            _callbacks=self.callbacks,
-        )
+            if not isinstance(self.predictor, yolo.yoloe.YOLOEVPDetectPredictor):
+                self.predictor = (predictor or yolo.yoloe.YOLOEVPDetectPredictor)(
+                    overrides={
+                        "task": self.model.task,
+                        "mode": "predict",
+                        "save": False,
+                        "verbose": refer_image is None,
+                        "batch": 1,
+                    },
+                    _callbacks=self.callbacks,
+                )
 
-        if len(visual_prompts):
             num_cls = (
                 max(len(set(c)) for c in visual_prompts["cls"])
                 if isinstance(source, list) and refer_image is None  # means multiple images
@@ -426,18 +426,19 @@ class YOLOE(Model):
             self.model.model[-1].nc = num_cls
             self.model.names = [f"object{i}" for i in range(num_cls)]
             self.predictor.set_prompts(visual_prompts.copy())
-
-        self.predictor.setup_model(model=self.model)
-
-        if refer_image is None and source is not None:
-            dataset = load_inference_source(source)
-            if dataset.mode in {"video", "stream"}:
-                # NOTE: set the first frame as refer image for videos/streams inference
-                refer_image = next(iter(dataset))[1][0]
-        if refer_image is not None and len(visual_prompts):
-            vpe = self.predictor.get_vpe(refer_image)
-            self.model.set_classes(self.model.names, vpe)
-            self.task = "segment" if isinstance(self.predictor, yolo.segment.SegmentationPredictor) else "detect"
-            self.predictor = None  # reset predictor
+            self.predictor.setup_model(model=self.model)
+
+            if refer_image is None and source is not None:
+                dataset = load_inference_source(source)
+                if dataset.mode in {"video", "stream"}:
+                    # NOTE: set the first frame as refer image for videos/streams inference
+                    refer_image = next(iter(dataset))[1][0]
+            if refer_image is not None:
+                vpe = self.predictor.get_vpe(refer_image)
+                self.model.set_classes(self.model.names, vpe)
+                self.task = "segment" if isinstance(self.predictor, yolo.segment.SegmentationPredictor) else "detect"
+                self.predictor = None  # reset predictor
+        elif isinstance(self.predictor, yolo.yoloe.YOLOEVPDetectPredictor):
+            self.predictor = None  # reset predictor if no visual prompts
 
         return super().predict(source, stream, **kwargs)

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/models/yolo/world/train.py

@@ -158,7 +158,7 @@ class WorldTrainer(DetectionTrainer):
                 return txt_map
         LOGGER.info(f"Caching text embeddings to '{cache_path}'")
         assert self.model is not None
-        txt_feats = self.model.get_text_pe(texts, batch, cache_clip_model=False)
+        txt_feats = de_parallel(self.model).get_text_pe(texts, batch, cache_clip_model=False)
         txt_map = dict(zip(texts, txt_feats.squeeze(0)))
         torch.save(txt_map, cache_path)
         return txt_map

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/models/yolo/yoloe/train.py

@@ -222,7 +222,7 @@ class YOLOETrainerFromScratch(YOLOETrainer, WorldTrainerFromScratch):
                 return txt_map
         LOGGER.info(f"Caching text embeddings to '{cache_path}'")
         assert self.model is not None
-        txt_feats = self.model.get_text_pe(texts, batch, without_reprta=True, cache_clip_model=False)
+        txt_feats = de_parallel(self.model).get_text_pe(texts, batch, without_reprta=True, cache_clip_model=False)
         txt_map = dict(zip(texts, txt_feats.squeeze(0)))
         torch.save(txt_map, cache_path)
         return txt_map

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/solutions/heatmap.py

@@ -124,6 +124,6 @@ class Heatmap(ObjectCounter):
             plot_im=plot_im,
             in_count=self.in_count,
             out_count=self.out_count,
-            classwise_count=dict(self.classwise_counts),
+            classwise_count=dict(self.classwise_count),
             total_tracks=len(self.track_ids),
         )

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/solutions/object_counter.py

@@ -43,7 +43,7 @@ class ObjectCounter(BaseSolution):
         self.in_count = 0  # Counter for objects moving inward
         self.out_count = 0  # Counter for objects moving outward
         self.counted_ids = []  # List of IDs of objects that have been counted
-        self.classwise_counts = defaultdict(lambda: {"IN": 0, "OUT": 0})  # Dictionary for counts, categorized by class
+        self.classwise_count = defaultdict(lambda: {"IN": 0, "OUT": 0})  # Dictionary for counts, categorized by class
         self.region_initialized = False  # Flag indicating whether the region has been initialized
 
         self.show_in = self.CFG["show_in"]
@@ -85,17 +85,17 @@ class ObjectCounter(BaseSolution):
                     # Vertical region: Compare x-coordinates to determine direction
                     if current_centroid[0] > prev_position[0]:  # Moving right
                         self.in_count += 1
-                        self.classwise_counts[self.names[cls]]["IN"] += 1
+                        self.classwise_count[self.names[cls]]["IN"] += 1
                     else:  # Moving left
                         self.out_count += 1
-                        self.classwise_counts[self.names[cls]]["OUT"] += 1
+                        self.classwise_count[self.names[cls]]["OUT"] += 1
                 # Horizontal region: Compare y-coordinates to determine direction
                 elif current_centroid[1] > prev_position[1]:  # Moving downward
                     self.in_count += 1
-                    self.classwise_counts[self.names[cls]]["IN"] += 1
+                    self.classwise_count[self.names[cls]]["IN"] += 1
                 else:  # Moving upward
                     self.out_count += 1
-                    self.classwise_counts[self.names[cls]]["OUT"] += 1
+                    self.classwise_count[self.names[cls]]["OUT"] += 1
                 self.counted_ids.append(track_id)
 
         elif len(self.region) > 2:  # Polygonal region
@@ -111,10 +111,10 @@ class ObjectCounter(BaseSolution):
                     and current_centroid[1] > prev_position[1]
                 ):  # Moving right or downward
                     self.in_count += 1
-                    self.classwise_counts[self.names[cls]]["IN"] += 1
+                    self.classwise_count[self.names[cls]]["IN"] += 1
                 else:  # Moving left or upward
                     self.out_count += 1
-                    self.classwise_counts[self.names[cls]]["OUT"] += 1
+                    self.classwise_count[self.names[cls]]["OUT"] += 1
                 self.counted_ids.append(track_id)
 
     def display_counts(self, plot_im) -> None:
@@ -132,7 +132,7 @@ class ObjectCounter(BaseSolution):
         labels_dict = {
             str.capitalize(key): f"{'IN ' + str(value['IN']) if self.show_in else ''} "
             f"{'OUT ' + str(value['OUT']) if self.show_out else ''}".strip()
-            for key, value in self.classwise_counts.items()
+            for key, value in self.classwise_count.items()
             if value["IN"] != 0 or value["OUT"] != 0 and (self.show_in or self.show_out)
         }
         if labels_dict:
@@ -190,6 +190,6 @@ class ObjectCounter(BaseSolution):
             plot_im=plot_im,
             in_count=self.in_count,
             out_count=self.out_count,
-            classwise_count=dict(self.classwise_counts),
+            classwise_count=dict(self.classwise_count),
             total_tracks=len(self.track_ids),
         )

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/solutions/solutions.py

@@ -808,10 +808,10 @@ class SolutionResults:
         filled_slots (int): The number of filled slots in a monitored area.
         email_sent (bool): A flag indicating whether an email notification was sent.
         total_tracks (int): The total number of tracked objects.
-        region_counts (Dict): The count of objects within a specific region.
+        region_counts (Dict[str, int]): The count of objects within a specific region.
         speed_dict (Dict[str, float]): A dictionary containing speed information for tracked objects.
         total_crop_objects (int): Total number of cropped objects using ObjectCropper class.
-        speed (Dict): Performance timing information for tracking and solution processing.
+        speed (Dict[str, float]): Performance timing information for tracking and solution processing.
     """
 
     def __init__(self, **kwargs):

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/utils/__init__.py

@@ -255,11 +255,8 @@ class DataExportMixin:
         Notes:
             Requires `lxml` package to be installed.
         """
-        from ultralytics.utils.checks import check_requirements
-
-        check_requirements("lxml")
         df = self.to_df(normalize=normalize, decimals=decimals)
-        return '<?xml version="1.0" encoding="utf-8"?>\n<root></root>' if df.empty else df.to_xml()
+        return '<?xml version="1.0" encoding="utf-8"?>\n<root></root>' if df.empty else df.to_xml(parser="etree")
 
     def to_html(self, normalize=False, decimals=5, index=False):
         """

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/utils/instance.py

@@ -406,6 +406,8 @@ class Instances:
                 | (self.keypoints[..., 1] < 0)
                 | (self.keypoints[..., 1] > h)
             ] = 0.0
+            self.keypoints[..., 0] = self.keypoints[..., 0].clip(0, w)
+            self.keypoints[..., 1] = self.keypoints[..., 1].clip(0, h)
 
     def remove_zero_area_boxes(self):
         """

{dgenerate_ultralytics_headless-8.3.159 → dgenerate_ultralytics_headless-8.3.160}/ultralytics/utils/metrics.py

@@ -1078,23 +1078,21 @@ class DetMetrics(SimpleClass, DataExportMixin):
            >>> detection_summary = results.summary()
            >>> print(detection_summary)
         """
-        scalars = {
-            "box-map": round(self.box.map, decimals),
-            "box-map50": round(self.box.map50, decimals),
-            "box-map75": round(self.box.map75, decimals),
-        }
         per_class = {
-            "box-p": self.box.p,
-            "box-r": self.box.r,
-            "box-f1": self.box.f1,
+            "Box-P": self.box.p,
+            "Box-R": self.box.r,
+            "Box-F1": self.box.f1,
         }
         return [
             {
-                "class_name": self.names[self.ap_class_index[i]],
+                "Class": self.names[self.ap_class_index[i]],
+                "Images": self.nt_per_image[self.ap_class_index[i]],
+                "Instances": self.nt_per_class[self.ap_class_index[i]],
                 **{k: round(v[i], decimals) for k, v in per_class.items()},
-                **scalars,
+                "mAP50": round(self.class_result(i)[2], decimals),
+                "mAP50-95": round(self.class_result(i)[3], decimals),
             }
-            for i in range(len(per_class["box-p"]))
+            for i in range(len(per_class["Box-P"]))
         ]
 
 
@@ -1213,19 +1211,14 @@ class SegmentMetrics(DetMetrics):
             >>> seg_summary = results.summary(decimals=4)
             >>> print(seg_summary)
         """
-        scalars = {
-            "mask-map": round(self.seg.map, decimals),
-            "mask-map50": round(self.seg.map50, decimals),
-            "mask-map75": round(self.seg.map75, decimals),
-        }
         per_class = {
-            "mask-p": self.seg.p,
-            "mask-r": self.seg.r,
-            "mask-f1": self.seg.f1,
+            "Mask-P": self.seg.p,
+            "Mask-R": self.seg.r,
+            "Mask-F1": self.seg.f1,
         }
         summary = DetMetrics.summary(self, normalize, decimals)  # get box summary
         for i, s in enumerate(summary):
-            s.update({**{k: round(v[i], decimals) for k, v in per_class.items()}, **scalars})
+            s.update({**{k: round(v[i], decimals) for k, v in per_class.items()}})
         return summary
 
 
@@ -1357,19 +1350,14 @@ class PoseMetrics(DetMetrics):
             >>> pose_summary = results.summary(decimals=4)
             >>> print(pose_summary)
         """
-        scalars = {
-            "pose-map": round(self.pose.map, decimals),
-            "pose-map50": round(self.pose.map50, decimals),
-            "pose-map75": round(self.pose.map75, decimals),
-        }
         per_class = {
-            "pose-p": self.pose.p,
-            "pose-r": self.pose.r,
-            "pose-f1": self.pose.f1,
+            "Pose-P": self.pose.p,
+            "Pose-R": self.pose.r,
+            "Pose-F1": self.pose.f1,
         }
         summary = DetMetrics.summary(self, normalize, decimals)  # get box summary
         for i, s in enumerate(summary):
-            s.update({**{k: round(v[i], decimals) for k, v in per_class.items()}, **scalars})
+            s.update({**{k: round(v[i], decimals) for k, v in per_class.items()}})
         return summary
 
 
@@ -1445,7 +1433,7 @@ class ClassifyMetrics(SimpleClass, DataExportMixin):
             >>> classify_summary = results.summary(decimals=4)
             >>> print(classify_summary)
         """
-        return [{"classify-top1": round(self.top1, decimals), "classify-top5": round(self.top5, decimals)}]
+        return [{"top1_acc": round(self.top1, decimals), "top5_acc": round(self.top5, decimals)}]
 
 
 class OBBMetrics(DetMetrics):