Create app.py

app.py

@@ -0,0 +1,39 @@
+import gradio as gr
+import tensorflow as tf
+import numpy as np
+import librosa
+
+# 1. Load your trained model (must be in the same folder, named model.h5)
+model = tf.keras.models.load_model("model.h5")
+
+# 2. Define labels & emojis (match your model’s output classes)
+EMOTIONS = ["Neutral", "Calm", "Happy", "Sad", "Angry", "Fearful", "Disgust", "Surprised"]
+EMOJI_MAP = {
+    "Neutral": "😐", "Calm": "😌", "Happy": "😄", "Sad": "😢",
+    "Angry": "😠", "Fearful": "😨", "Disgust": "🤢", "Surprised": "😲"
+}
+
+def predict_emotion(audio_path):
+    # Load & preprocess audio
+    y, sr = librosa.load(audio_path, sr=22050)
+    mfcc = librosa.feature.mfcc(y=y, sr=sr, n_mfcc=40)
+    features = np.mean(mfcc.T, axis=0).reshape(1, -1)
+    
+    # Run model
+    preds = model.predict(features)
+    idx = np.argmax(preds, axis=1)[0]
+    label = EMOTIONS[idx]
+    emoji = EMOJI_MAP[label]
+    return f"{label} {emoji}"
+
+# 3. Build Gradio Interface
+demo = gr.Interface(
+    fn=predict_emotion,
+    inputs=gr.Audio(source="upload", type="filepath", label="Upload a .wav file"),
+    outputs=gr.Text(label="Predicted Emotion"),
+    title="🎤 Voice Emotion AI",
+    description="Upload a voice clip (.wav) to detect the speaker’s emotion."
+)
+
+if __name__ == "__main__":
+    demo.launch()