Update app.py

app.py

@@ -2,7 +2,6 @@ import gradio as gr
 import tensorflow as tf
 import librosa
 import numpy as np
-import tempfile
 
 # Diccionario de etiquetas
 labels = ['down', 'go', 'left', 'no', 'off', 'on', 'right', 'stop', 'up', 'yes']
@@ -21,33 +20,23 @@ def extract_features(file_name):
     return mfccsscaled
 
 def classify_audio(audio_file):
-    print(f"Tipo de audio_file: {type(audio_file)}") 
+    print(f"Tipo de audio_file: {type(audio_file)}")  # Debería imprimir <class 'str'>
 
-    with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as tmp_file:
-        tmp_file.write(audio_file.read())
-        tmp_file_path = tmp_file.name
+    # Preprocesa el audio directamente
+    features = extract_features(audio_file)
 
-    # Preprocesa el audio (con extract_features())
-    features = extract_features(tmp_file_path)
-
-    # Si features es None, hubo un error en extract_features
     if features is None:
         return "Error al procesar el audio" 
     
-    features = features.reshape(1, -1) # Redimensiona a (1, 40)
-    
-    # Si tu modelo necesita 3 dimensiones, redimensiona a (1, 40, 1)
-    # features = features.reshape(1, 40, 1)
+    features = features.reshape(1, -1)
 
-    # Carga del modelo para usar la CPU
+    # Carga el modelo (asegúrate que 'my_model.h5' esté en el mismo directorio)
     model = tf.keras.models.load_model('my_model.h5', compile=False)
 
-    # Realiza la predicción en la CPU
     with tf.device('/CPU:0'):
         prediction = model.predict(features)
-        predicted_label_index = np.argmax(prediction) 
+        predicted_label_index = np.argmax(prediction)
     
-    # Devuelve la etiqueta predicha
     predicted_label = labels[predicted_label_index]
     return predicted_label