Update app.py

app.py

@@ -1,27 +1,27 @@
 import gradio as gr
-from transformers import AutoModelForVideoClassification, AutoTokenizer, VideoClassificationPipeline
+from transformers import VideoMAEForVideoClassification, VideoMAEFeatureExtractor, VideoClassificationPipeline
 import cv2  # OpenCV for video processing
 
 # Model ID for video classification (UCF101 subset)
-model_id = "sayakpaul/videomae-base-finetuned-ucf101-subset"
+model_id = "MCG-NJU/videomae-base"
 
 def analyze_video(video):
     # Extract key frames from the video using OpenCV
     frames = extract_key_frames(video)
     
-    # Load model and tokenizer manually
-    model = AutoModelForVideoClassification.from_pretrained(model_id)
-    tokenizer = AutoTokenizer.from_pretrained(model_id)
+    # Load model and feature extractor manually
+    model = VideoMAEForVideoClassification.from_pretrained(model_id)
+    feature_extractor = VideoMAEFeatureExtractor.from_pretrained(model_id)
 
     # Create the pipeline
-    classifier = VideoClassificationPipeline(model=model, tokenizer=tokenizer, device=-1)
+    classifier = VideoClassificationPipeline(model=model, feature_extractor=feature_extractor, device=-1)
     
     # Analyze key frames using video classification model
     results = []
     for frame in frames:
         # OpenCV uses BGR, convert to RGB for the model
         frame_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
-        predictions = classifier(images=[frame_rgb])  # Assuming model outputs probabilities
+        predictions = classifier([frame_rgb])  # Assuming model outputs probabilities
         # Analyze predictions for insights related to the play
         result = analyze_predictions_ucf101(predictions)
         results.append(result)