update

ytracking/ultralytics/models/nas/__init__.py

@@ -0,0 +1,7 @@
+# Ultralytics YOLO 🚀, AGPL-3.0 license
+
+from .model import NAS
+from .predict import NASPredictor
+from .val import NASValidator
+
+__all__ = 'NASPredictor', 'NASValidator', 'NAS'

ytracking/ultralytics/models/nas/model.py

@@ -0,0 +1,61 @@
+# Ultralytics YOLO 🚀, AGPL-3.0 license
+"""
+YOLO-NAS model interface.
+
+Example:
+    ```python
+    from ultralytics import NAS
+
+    model = NAS('yolo_nas_s')
+    results = model.predict('ultralytics/assets/bus.jpg')
+    ```
+"""
+
+from pathlib import Path
+
+import torch
+
+from ultralytics.engine.model import Model
+from ultralytics.utils.torch_utils import model_info, smart_inference_mode
+
+from .predict import NASPredictor
+from .val import NASValidator
+
+
+class NAS(Model):
+
+    def __init__(self, model='yolo_nas_s.pt') -> None:
+        assert Path(model).suffix not in ('.yaml', '.yml'), 'YOLO-NAS models only support pre-trained models.'
+        super().__init__(model, task='detect')
+
+    @smart_inference_mode()
+    def _load(self, weights: str, task: str):
+        # Load or create new NAS model
+        import super_gradients
+        suffix = Path(weights).suffix
+        if suffix == '.pt':
+            self.model = torch.load(weights)
+        elif suffix == '':
+            self.model = super_gradients.training.models.get(weights, pretrained_weights='coco')
+        # Standardize model
+        self.model.fuse = lambda verbose=True: self.model
+        self.model.stride = torch.tensor([32])
+        self.model.names = dict(enumerate(self.model._class_names))
+        self.model.is_fused = lambda: False  # for info()
+        self.model.yaml = {}  # for info()
+        self.model.pt_path = weights  # for export()
+        self.model.task = 'detect'  # for export()
+
+    def info(self, detailed=False, verbose=True):
+        """
+        Logs model info.
+
+        Args:
+            detailed (bool): Show detailed information about model.
+            verbose (bool): Controls verbosity.
+        """
+        return model_info(self.model, detailed=detailed, verbose=verbose, imgsz=640)
+
+    @property
+    def task_map(self):
+        return {'detect': {'predictor': NASPredictor, 'validator': NASValidator}}

ytracking/ultralytics/models/nas/predict.py

@@ -0,0 +1,35 @@
+# Ultralytics YOLO 🚀, AGPL-3.0 license
+
+import torch
+
+from ultralytics.engine.predictor import BasePredictor
+from ultralytics.engine.results import Results
+from ultralytics.utils import ops
+
+
+class NASPredictor(BasePredictor):
+
+    def postprocess(self, preds_in, img, orig_imgs):
+        """Postprocess predictions and returns a list of Results objects."""
+
+        # Cat boxes and class scores
+        boxes = ops.xyxy2xywh(preds_in[0][0])
+        preds = torch.cat((boxes, preds_in[0][1]), -1).permute(0, 2, 1)
+
+        preds = ops.non_max_suppression(preds,
+                                        self.args.conf,
+                                        self.args.iou,
+                                        agnostic=self.args.agnostic_nms,
+                                        max_det=self.args.max_det,
+                                        classes=self.args.classes)
+
+        if not isinstance(orig_imgs, list):  # input images are a torch.Tensor, not a list
+            orig_imgs = ops.convert_torch2numpy_batch(orig_imgs)
+
+        results = []
+        for i, pred in enumerate(preds):
+            orig_img = orig_imgs[i]
+            pred[:, :4] = ops.scale_boxes(img.shape[2:], pred[:, :4], orig_img.shape)
+            img_path = self.batch[0][i]
+            results.append(Results(orig_img, path=img_path, names=self.model.names, boxes=pred))
+        return results

ytracking/ultralytics/models/nas/val.py

@@ -0,0 +1,24 @@
+# Ultralytics YOLO 🚀, AGPL-3.0 license
+
+import torch
+
+from ultralytics.models.yolo.detect import DetectionValidator
+from ultralytics.utils import ops
+
+__all__ = ['NASValidator']
+
+
+class NASValidator(DetectionValidator):
+
+    def postprocess(self, preds_in):
+        """Apply Non-maximum suppression to prediction outputs."""
+        boxes = ops.xyxy2xywh(preds_in[0][0])
+        preds = torch.cat((boxes, preds_in[0][1]), -1).permute(0, 2, 1)
+        return ops.non_max_suppression(preds,
+                                       self.args.conf,
+                                       self.args.iou,
+                                       labels=self.lb,
+                                       multi_label=False,
+                                       agnostic=self.args.single_cls,
+                                       max_det=self.args.max_det,
+                                       max_time_img=0.5)