Initial commit with FastAPI + Gradio app

app.py

@@ -1,9 +1,9 @@
 import os
-import shutil
 import torchaudio
 import gradio as gr
+from fastapi import FastAPI, HTTPException, File, UploadFile
 from speechbrain.inference import SpeakerRecognition
-from fastapi import HTTPException
+from fastapi.responses import JSONResponse
 
 # Initialize the speaker verification model
 speaker_verification = SpeakerRecognition.from_hparams(
@@ -53,16 +53,52 @@ def compare_voices(file1, file2):
 
     return {"Similarity Score": f"{score:.4f}", "Same User Prediction": is_same_user}
 
-# Create Gradio Interface for the API
-api = gr.Interface(
-    fn=compare_voices,
-    inputs=[
+# FastAPI app
+app = FastAPI()
+
+@app.post("/compare_voices/")
+async def compare_voices_api(file1: UploadFile = File(...), file2: UploadFile = File(...)):
+    """
+    Compare two audio files and return the similarity score and prediction.
+    """
+    # Save uploaded files temporarily
+    file1_path = os.path.join(UPLOAD_FOLDER, file1.filename)
+    file2_path = os.path.join(UPLOAD_FOLDER, file2.filename)
+
+    with open(file1_path, "wb") as f1:
+        f1.write(await file1.read())
+    with open(file2_path, "wb") as f2:
+        f2.write(await file2.read())
+
+    # Get similarity score
+    score, is_same_user = get_similarity(file1_path, file2_path)
+
+    if is_same_user is None:
+        raise HTTPException(status_code=500, detail="Error in processing files: " + score)
+
+    return JSONResponse(content={"Similarity Score": f"{score:.4f}", "Same User Prediction": is_same_user})
+
+# Gradio interface function
+def gradio_interface():
+    return gr.Interface(
+        fn=compare_voices,
+        inputs=[
+            gr.Audio(source="upload", type="file", label="First Audio File"),
+            gr.Audio(source="upload", type="file", label="Second Audio File")
+        ],
+        outputs="json",  # Output results as JSON
+        live=False  # No live interface, just the API
+    )
+
+# Launch Gradio as a web interface
+@app.on_event("startup")
+async def startup():
+    gr.Interface(fn=compare_voices, inputs=[
         gr.Audio(source="upload", type="file", label="First Audio File"),
         gr.Audio(source="upload", type="file", label="Second Audio File")
-    ],
-    outputs="json",  # Output results as JSON
-    live=False  # No live interface, just the API
-)
+    ], outputs="json", live=False).launch(share=True, inline=True)
 
-# Launch the API as an HTTP server
-api.queue().launch(server_name="0.0.0.0", server_port=8080, share=False)
+# Launch FastAPI with Gradio and FastAPI routes
+if __name__ == "__main__":
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=5000)

requirements.txt

@@ -1,9 +1 @@
-fastapi
-uvicorn
-numpy
-scikit-learn
-joblib
-torchaudio
-speechbrain
-python-multipart
-gradio
+git+https://github.com/speechbrain/speechbrain.git@develop