Update video_processing.py

video_processing.py

@@ -154,9 +154,11 @@ def process_video(video_path, anomaly_threshold, desired_fps, progress=None):
         face_samples = get_all_face_samples(organized_faces_folder, output_folder, largest_cluster)
 
         progress(0.8, "Extracting audio and performing voice analysis")
+        
         audio_path = extract_audio_from_video(video_path)
         diarization = diarize_speakers(audio_path)
-        voice_embeddings = get_speaker_embeddings(audio_path, diarization, "pyannote/embedding")
+        voice_embeddings, audio_duration = get_speaker_embeddings(audio_path, diarization, "pyannote/embedding")
+
 
         progress(0.85, "Performing anomaly detection")
         embedding_columns = [col for col in df.columns if col.startswith('Raw_Embedding_')]
@@ -170,7 +172,9 @@ def process_video(video_path, anomaly_threshold, desired_fps, progress=None):
             if len(X_posture) == 0:
                 raise ValueError("No valid posture data found")
 
-            X_voice = np.array([emb['embedding'] for emb in voice_embeddings])
+            aligned_voice_embeddings = align_voice_embeddings(voice_embeddings, frame_count, original_fps, audio_duration)
+            
+            X_voice = np.array([emb for emb in aligned_voice_embeddings])
 
             mse_embeddings, mse_posture, mse_voice = anomaly_detection(X_embeddings, X_posture, X_voice)