| import gradio as gr |
| import numpy as np |
| from audio_features import extract_features_with_time_series |
| from model_utils import load_model, prepare_input_for_model, interpret_prediction |
|
|
| MODEL_PATH = "best_audio_model.keras" |
|
|
| |
| model = load_model(MODEL_PATH) |
|
|
| def predict_audio(file_path): |
| try: |
| features = extract_features_with_time_series(file_path) |
| if features is None: |
| return {"Error": "Feature extraction failed. Please check the audio file."} |
|
|
| input_data = prepare_input_for_model(features, model) |
| raw_pred = model.predict(input_data) |
| result = interpret_prediction(raw_pred) |
| return result |
| except Exception as e: |
| return {"Error": str(e)} |
|
|
| iface = gr.Interface( |
| fn=predict_audio, |
| inputs=gr.Audio(type="filepath", label="Upload an audio file"), |
| outputs=gr.Label(num_top_classes=2, label="Prediction"), |
| title="Audio Deepfake Detection", |
| description="Upload an audio clip to check if it's Real or Fake using your trained .keras model." |
| ) |
|
|
| if __name__ == "__main__": |
| iface.launch(server_name="0.0.0.0", server_port=7860) |
|
|
