Upload app.py

app.py

@@ -0,0 +1,61 @@
+import gradio as gr
+import numpy as np
+import tensorflow as tf
+import librosa
+import librosa.util
+
+# Define your predict_class function
+def predict_class(file_path, model, labels):
+    # Extract MFCC features
+    y, sr = librosa.load(file_path, sr=None)
+    mfcc = librosa.feature.mfcc(y=y, sr=sr, n_mfcc=13)
+    
+    # Pad or truncate to 100 frames along axis 1
+    mfcc = librosa.util.fix_length(mfcc, size=100, axis=1)
+
+    # Ensure mfcc has shape (13, 100)
+    if mfcc.shape[0] != 13:
+        mfcc = librosa.util.fix_length(mfcc, size=13, axis=0)
+
+    # Add batch and channel dimensions
+    mfcc = mfcc[np.newaxis, ..., np.newaxis]  # Shape: (1, 13, 100, 1)
+
+    # Predict using the model
+    prediction = model.predict(mfcc)
+    predicted_class = labels[np.argmax(prediction)]
+    return predicted_class
+
+# Load your pre-trained model
+model = tf.keras.models.load_model("voice_classification_modelm.h5")
+
+# Define the class labels based on your folder names
+labels = [
+    "all_vowels_healthy",
+    "allvowels_functional",
+    "allvowels_laryngitis",
+    "allvowels_lukoplakia",
+    "allvowels_psychogenic",
+    "allvowels_rlnp",
+    "allvowels_sd"
+]
+
+# Define the Gradio function
+def classify_audio(audio_file):
+    try:
+        predicted_class = predict_class(audio_file, model, labels)
+        return f"Predicted Class: {predicted_class}"
+    except Exception as e:
+        return f"Error: {str(e)}"
+
+# Create the Gradio interface
+interface = gr.Interface(
+    fn=classify_audio,
+    inputs=gr.Audio(source="upload", type="filepath", label="Upload an Audio File"),
+    outputs=gr.Textbox(label="Predicted Class"),
+    title="Voice Classification",
+    description="Upload an audio file to classify its voice type.",
+    examples=["example_audio.wav"]  # Replace with paths to sample audio files
+)
+
+# Launch the app
+interface.launch()