Update app.py

app.py

@@ -84,6 +84,31 @@ def process_audio(audio_file, max_duration=600):  # limit to 5 minutes initially
         st.error(f"Error processing audio: {str(e)}")
         return None
 
+def format_speaker_segments(diarization_result):
+    """Process and format speaker segments by removing very short segments and merging consecutive ones"""
+    formatted_segments = []
+    min_duration = 0.3  # Minimum duration threshold in seconds
+    
+    for turn, _, speaker in diarization_result.itertracks(yield_label=True):
+        duration = turn.end - turn.start
+        
+        # Skip very short segments
+        if duration < min_duration:
+            continue
+            
+        # Add segment if it's the first one or from a different speaker
+        if not formatted_segments or formatted_segments[-1]['speaker'] != speaker:
+            formatted_segments.append({
+                'speaker': speaker,
+                'start': turn.start,
+                'end': turn.end
+            })
+        # Extend the end time if it's the same speaker
+        else:
+            formatted_segments[-1]['end'] = turn.end
+    
+    return formatted_segments
+
 def main():
     st.title("Multi-Speaker Audio Analyzer")
     st.write("Upload an audio file (MP3/WAV) up to 5 minutes long for best performance")
@@ -92,7 +117,7 @@ def main():
 
     if uploaded_file:
         # Display file info
-        file_size = len(uploaded_file.getvalue()) / (1024 * 1024)  # Convert to MB
+        file_size = len(uploaded_file.getvalue()) / (1024 * 1024)
         st.write(f"File size: {file_size:.2f} MB")
         
         # Display audio player
@@ -105,14 +130,44 @@ def main():
                 results = process_audio(uploaded_file)
                 
                 if results:
-                    # Display results in tabs
                     tab1, tab2, tab3 = st.tabs(["Speakers", "Transcription", "Summary"])
                     
                     with tab1:
-                        st.write("Speaker Segments:")
-                        for turn, _, speaker in results["diarization"].itertracks(yield_label=True):
-                            st.write(f"{speaker}: {turn.start:.1f}s → {turn.end:.1f}s")
+                        st.write("Speaker Timeline:")
+                        
+                        # Process speaker segments
+                        segments = format_speaker_segments(results["diarization"])
+                        
+                        # Display segments in a more organized way
+                        for segment in segments:
+                            # Create columns for better layout
+                            col1, col2, col3 = st.columns([2,1,6])
+                            
+                            with col1:
+                                # Show speaker with consistent color
+                                speaker_num = int(segment['speaker'].split('_')[1])
+                                colors = ['🔵', '🔴', '🟢', '🟡', '🟣']  # Different colors for different speakers
+                                speaker_color = colors[speaker_num % len(colors)]
+                                st.write(f"{speaker_color} {segment['speaker']}")
+                            
+                            with col2:
+                                # Format time more cleanly
+                                start_time = f"{int(segment['start']):02d}:{(segment['start']%60):04.1f}"
+                                end_time = f"{int(segment['end']):02d}:{(segment['end']%60):04.1f}"
+                                st.write(f"{start_time} →")
+                            
+                            with col3:
+                                st.write(f"{end_time}")
+                            
+                            # Add a small separator
+                            st.markdown("---")
+                        
+                        # Add legend
+                        st.write("\nSpeaker Legend:")
+                        for i in range(len(set(s['speaker'] for s in segments))):
+                            st.write(f"{colors[i]} SPEAKER_{i:02d}")
                     
+                    # Keep original transcription and summary tabs
                     with tab2:
                         st.write("Transcription:")
                         st.write(results["transcription"])