Remove python related files

app.py

@@ -25,19 +25,20 @@ from utils import cfg, load_config, load_onnx_model
 
 # Configuration constants
 CHECKPOINTS = [
-    "ustc-community/dfine-medium-obj2coco",
-    "ustc-community/dfine-medium-coco",    
-    "ustc-community/dfine-medium-obj365",
-    "ustc-community/dfine-nano-coco",
-    "ustc-community/dfine-small-coco",
-    "ustc-community/dfine-large-coco",
-    "ustc-community/dfine-xlarge-coco",
-    "ustc-community/dfine-small-obj365",
-    "ustc-community/dfine-large-obj365",
-    "ustc-community/dfine-xlarge-obj365",
-    "ustc-community/dfine-small-obj2coco",
-    "ustc-community/dfine-large-obj2coco-e25",
-    "ustc-community/dfine-xlarge-obj2coco",
+    "yolo8n-cpu",
+    # "ustc-community/dfine-medium-obj2coco",
+    # "ustc-community/dfine-medium-coco",    
+    # "ustc-community/dfine-medium-obj365",
+    # "ustc-community/dfine-nano-coco",
+    # "ustc-community/dfine-small-coco",
+    # "ustc-community/dfine-large-coco",
+    # "ustc-community/dfine-xlarge-coco",
+    # "ustc-community/dfine-small-obj365",
+    # "ustc-community/dfine-large-obj365",
+    # "ustc-community/dfine-xlarge-obj365",
+    # "ustc-community/dfine-small-obj2coco",
+    # "ustc-community/dfine-large-obj2coco-e25",
+    # "ustc-community/dfine-xlarge-obj2coco",
 ]
 DEFAULT_CHECKPOINT = CHECKPOINTS[0]
 DEFAULT_CONFIDENCE_THRESHOLD = 0.3
@@ -109,9 +110,13 @@ def detect_objects(
     classes: Optional[List[str]] = None,
 ):
 
-    device = "cuda" if torch.cuda.is_available() else "cpu"
-    model, image_processor = get_model_and_processor(checkpoint)
-    model = model.to(device)
+    # device = "cuda" if torch.cuda.is_available() else "cpu"
+    # model, image_processor = get_model_and_processor(checkpoint)
+    # model = model.to(device)
+    
+    load_config(cfg, f'configs/{checkpoint}.yaml')
+    pipeline = build_pipeline(cfg.pipeline)
+    load_onnx_model(pipeline.detector, 'downloads/yolo8n-416.onnx')
 
     if classes is not None:
         wrong_classes = [cls for cls in classes if cls not in model.config.label2id]
@@ -130,22 +135,35 @@ def detect_objects(
     for batch in tqdm.tqdm(batches, desc="Processing frames"):
 
         # preprocess images
-        inputs = image_processor(images=batch, return_tensors="pt")
-        inputs = inputs.to(device).to(TORCH_DTYPE)
+        # inputs = image_processor(images=batch, return_tensors="pt")
+        # inputs = inputs.to(device).to(TORCH_DTYPE)
 
-        # forward pass
-        with torch.no_grad():
-            outputs = model(**inputs)
+        # # forward pass
+        # with torch.no_grad():
+        #     outputs = model(**inputs)
+        
+        outputs = []
+        for i in range(len(batch)):
+            img = batch[i]
+            output_ = pipeline(img)
+            output_1 = {
+            "scores": torch.from_numpy(output_.confidence) if isinstance(output_.confidence, np.ndarray) else output_.confidence,
+            "labels": torch.from_numpy(output_.class_id) if isinstance(output_.class_id, np.ndarray) else output_.class_id,
+            "boxes": torch.from_numpy(output_.xyxy) if isinstance(output_.xyxy, np.ndarray) else output_.xyxy,
+            }
+            outputs.append(output_1)
+        
 
         # postprocess outputs
-        if target_size:
-            target_sizes = [target_size] * len(batch)
-        else:
-            target_sizes = [(image.shape[0], image.shape[1]) for image in batch]
+        # if target_size:
+        #     target_sizes = [target_size] * len(batch)
+        # else:
+        #     target_sizes = [(image.shape[0], image.shape[1]) for image in batch]
 
-        batch_results = image_processor.post_process_object_detection(
-            outputs, target_sizes=target_sizes, threshold=confidence_threshold
-        )
+        # batch_results = image_processor.post_process_object_detection(
+        #     outputs, target_sizes=target_sizes, threshold=confidence_threshold
+        # )
+        batch_results = outputs
 
         results.extend(batch_results)
 
@@ -156,7 +174,8 @@ def detect_objects(
             keep = torch.isin(results[i]["labels"], torch.tensor(keep_ids))
             results[i] = {k: v[keep] for k, v in results[i].items()}
     
-    return results, model.config.id2label
+    # return results, model.config.id2label
+    return results, pipeline.detector.get_category_mapping()
 
 
 def process_image(
@@ -566,7 +585,4 @@ with gr.Blocks(theme=gr.themes.Ocean()) as demo:
     )
 
 if __name__ == "__main__":
-    load_config(cfg, 'configs/yolo8n-bytetrack-cpu.yaml')
-    pipeline = build_pipeline(cfg.pipeline)
-    load_onnx_model(pipeline.detector, 'downloads/yolo8n-416.onnx')
     demo.queue(max_size=20).launch()

configs/{yolo8n-bytetrack-cpu.yaml → yolo8n-cpu.yaml}

@@ -11,4 +11,5 @@ pipeline:
     device: cpu
 
   tracker:
-    algorithm: bytetrack
+    algorithm: dummytrack
+    

pipeline/trackers/__init__.py

@@ -1,8 +1,14 @@
 from .bytetrack import ByteTrack
+from .dummytrack import DummyTrack
 
-def build_tracker(cfg):
+def build_tracker(config):
     """
     Build the tracking model based on the provided configuration.
     """
     # Initialize the tracker
-    return ByteTrack()
+    if config.algorithm == "bytetrack":
+        return ByteTrack()
+    elif config.algorithm == "dummytrack":
+        return DummyTrack()
+    else:
+        raise ValueError(f"Unknown tracker type: {config.tracker.algorithm}")

pipeline/trackers/dummytrack.py

@@ -0,0 +1,7 @@
+class DummyTrack:
+    def __init__(self, *args, **kwargs):
+        pass
+
+    def __call__(self, detections, *args, **kwargs):
+        # Just return the detections as-is
+        return detections