feat: add YoloV7 Tiny support

app.py

@@ -6,14 +6,14 @@ from inference import inference
 
 input_video = gr.Video(mirror_webcam=False)
 
-dd_model = gr.Dropdown(choices=["YoloV7"], value="YoloV7", label="Model")
+dd_model = gr.Dropdown(choices=["YoloV7", "YoloV7 Tiny"], value="YoloV7", label="Model")
 
 cb_motion_estimation = gr.Checkbox(value=True, label="Track camera movement")
 
 cb_path_draw = gr.Checkbox(value=True, label="Draw objects paths")
 
 dd_track_points = gr.Dropdown(
-    choices=["Boxes", "Centroid"], value="Boxes", label="Detections style"
+    choices=["Bounding box", "Centroid"], value="Bounding box", label="Detections style"
 )
 
 slide_threshold = gr.Slider(minimum=0, maximum=1, value=0.25, label="Model confidence threshold")

inference.py

@@ -8,13 +8,13 @@ from norfair.camera_motion import HomographyTransformationGetter, MotionEstimato
 
 from inference_utils import (
     YOLO,
-    ModelsPath,
-    Style,
     center,
     clean_videos,
     draw,
     euclidean_distance,
     iou,
+    models_path,
+    style,
     yolo_detections_to_norfair_detections,
 )
 
@@ -22,16 +22,6 @@ DISTANCE_THRESHOLD_BBOX: float = 3.33
 DISTANCE_THRESHOLD_CENTROID: int = 30
 MAX_DISTANCE: int = 10000
 
-parser = argparse.ArgumentParser(description="Track objects in a video.")
-parser.add_argument("--img-size", type=int, default="720", help="YOLOv7 inference size (pixels)")
-parser.add_argument(
-    "--iou-threshold", type=float, default="0.45", help="YOLOv7 IOU threshold for NMS"
-)
-parser.add_argument(
-    "--classes", nargs="+", type=int, help="Filter by class: --classes 0, or --classes 0 2 3"
-)
-args = parser.parse_args()
-
 
 def inference(
     input_video: str,
@@ -48,8 +38,8 @@ def inference(
     coord_transformations = None
     paths_drawer = None
     fix_paths = False
-    track_points = Style[track_points].value
-    model = YOLO(ModelsPath[model].value)
+    track_points = style[track_points]
+    model = YOLO(models_path[model])
     video = Video(input_path=input_video, output_path=output_path)
 
     if motion_estimation and drawing_paths:
@@ -79,11 +69,7 @@ def inference(
 
     for frame in video:
         yolo_detections = model(
-            frame,
-            conf_threshold=model_threshold,
-            iou_threshold=args.iou_threshold,
-            image_size=720,
-            classes=args.classes,
+            frame, conf_threshold=model_threshold, iou_threshold=0.45, image_size=720
         )
 
         mask = np.ones(frame.shape[:2], frame.dtype)

inference_utils.py

@@ -15,13 +15,9 @@ DISTANCE_THRESHOLD_CENTROID: int = 30
 MAX_DISTANCE: int = 10000
 
 
-class ModelsPath(Enum):
-    YoloV7 = "models/yolov7.pt"
+models_path = {"YoloV7": "models/yolov7.pt", "YoloV7 Tiny": "models/yolov7-tiny.pt"}
 
-
-class Style(Enum):
-    Boxes = "bbox"
-    Centroid = "centroid"
+style = {"Bounding box": "bbox", "Centroid": "centroid"}
 
 
 class YOLO: