Update app.py

app.py

@@ -6,9 +6,16 @@ import cv2  # OpenCV for video processing
 # Model ID for video classification (UCF101 subset)
 model_id = "MCG-NJU/videomae-base"
 
+# Parameters for frame extraction
+TARGET_FRAME_COUNT = 16
+FRAME_SIZE = (224, 224)  # Expected frame size for the model
+
 def analyze_video(video):
     # Extract key frames from the video using OpenCV
-    frames = extract_key_frames(video)
+    frames = extract_key_frames(video, TARGET_FRAME_COUNT)
+    
+    # Resize frames to the expected size
+    frames = [cv2.resize(frame, FRAME_SIZE) for frame in frames]
 
     # Load model and feature extractor manually
     model = VideoMAEForVideoClassification.from_pretrained(model_id)
@@ -35,16 +42,21 @@ def analyze_video(video):
 
     return final_result
 
-def extract_key_frames(video):
+def extract_key_frames(video, target_frame_count):
     cap = cv2.VideoCapture(video)
     frames = []
     frame_count = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
-    fps = int(cap.get(cv2.CAP_PROP_FPS))
     
-    for i in range(frame_count):
+    # Calculate interval for frame extraction
+    interval = max(1, frame_count // target_frame_count)
+    
+    for i in range(0, frame_count, interval):
+        cap.set(cv2.CAP_PROP_POS_FRAMES, i)
         ret, frame = cap.read()
-        if ret and i % (fps // 2) == 0:  # Extract a frame every half second
+        if ret:
             frames.append(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB))  # Convert to RGB
+        if len(frames) >= target_frame_count:
+            break
     
     cap.release()
     return frames