Initial speaker diarization app with pyannote 3.1

README.md

@@ -1,12 +1,16 @@
 ---
 title: Speaker Diarization Pyannote
-emoji: 💻
-colorFrom: gray
-colorTo: red
+emoji: 🎤
+colorFrom: blue
+colorTo: purple
 sdk: gradio
-sdk_version: 5.32.0
+sdk_version: 4.20.0
 app_file: app.py
 pinned: false
+license: mit
+hardware: t4-small
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Speaker Diarization with Pyannote
+
+This space performs speaker diarization using pyannote.audio 3.1

app.py

@@ -0,0 +1,56 @@
+import gradio as gr
+from pyannote.audio import Pipeline
+import torch
+import torchaudio
+from huggingface_hub import login
+import os
+
+# Login to Hugging Face if token is available
+hf_token = os.environ.get("HF_TOKEN")
+if hf_token:
+    login(token=hf_token)
+
+# Initialize the pipeline
+pipeline = Pipeline.from_pretrained(
+    "pyannote/speaker-diarization-3.1",
+    use_auth_token=hf_token
+)
+
+# Send pipeline to GPU if available
+if torch.cuda.is_available():
+    pipeline.to(torch.device("cuda"))
+
+def diarize_audio(audio_file):
+    """Process audio file and return diarization results"""
+    try:
+        # Apply pretrained pipeline
+        diarization = pipeline(audio_file)
+        
+        # Format results
+        results = []
+        for turn, _, speaker in diarization.itertracks(yield_label=True):
+            results.append(
+                f"Speaker {speaker}: {turn.start:.1f}s - {turn.end:.1f}s"
+            )
+        
+        if not results:
+            return "No speakers detected in the audio."
+        
+        return "\n".join(results)
+        
+    except Exception as e:
+        return f"Error processing audio: {str(e)}"
+
+# Create Gradio interface
+demo = gr.Interface(
+    fn=diarize_audio,
+    inputs=gr.Audio(type="filepath", label="Upload Audio File"),
+    outputs=gr.Textbox(label="Diarization Results", lines=10),
+    title="Speaker Diarization with Pyannote 3.1",
+    description="Upload an audio file to identify different speakers and their speaking times.",
+    examples=[],
+    cache_examples=False
+)
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,5 @@
+pyannote.audio==3.1.1
+torch==2.0.1
+torchaudio==2.0.2
+gradio==4.20.0
+huggingface_hub