newMonday

app.py

@@ -34,11 +34,25 @@ def diarize_audio(audio_path):
     diarization = diarization_pipeline(audio_path)
     speaker_segments = {}
 
+    # Ajustement : fusionner les segments proches si le silence est trop court
+    min_silence_duration = 0.5  # Durée minimale du silence pour être considéré comme un changement de speaker
+    previous_speaker = None
+    last_end = 0
+
     for turn, _, speaker in diarization.itertracks(yield_label=True):
         start, end = turn.start, turn.end
-        if speaker not in speaker_segments:
-            speaker_segments[speaker] = []
-        speaker_segments[speaker].append((start, end))
+        
+        # Si le silence entre deux segments est trop court, on fusionne avec le speaker précédent
+        if previous_speaker is not None and start - last_end < min_silence_duration:
+            speaker_segments[previous_speaker].append((start, end))
+        else:
+            # Nouveau speaker ou silence long : on l'ajoute comme un segment distinct
+            if speaker not in speaker_segments:
+                speaker_segments[speaker] = []
+            speaker_segments[speaker].append((start, end))
+        
+        previous_speaker = speaker
+        last_end = end
 
     return speaker_segments