Update app.py

app.py

@@ -224,7 +224,8 @@ def transcribe_video_with_speakers(video_path):
         speaker = segment["speaker"]
         if speaker not in speaker_audio:
             speaker_audio[speaker] = []
-        speaker_audio[speaker].append((segment["start"], segment["end"]))
+        if end > start and (end - start) > 0.05:  # Require >50ms duration
+            speaker_audio[speaker].append((segment["start"], segment["end"]))
 
     # Collapse and truncate speaker audio
     speaker_sample_paths = {}