Update app.py

app.py

@@ -3,20 +3,20 @@ import tensorflow as tf
 import numpy as np
 
 IMG_SIZE = (224, 224)
+MODEL_PATH = "dental_classifier_model.keras"
 CLASS_NAMES = ['no_valido', 'valido']
 
+# Cargar modelo una vez al inicio
+model = tf.keras.models.load_model(MODEL_PATH)
+
 def preprocess_image(img):
+    # Redimensionar y normalizar la imagen
     img = tf.image.resize(img, IMG_SIZE)
     img_array = tf.expand_dims(img, 0)
     img_array = img_array / 255.0
     return img_array
 
-def predict(rx_image, keras_model_file):
-    model_path = "temp_model.keras"
-    with open(model_path, "wb") as f:
-        f.write(keras_model_file.read())
-    
-    model = tf.keras.models.load_model(model_path)
+def predict(rx_image):
     img_array = preprocess_image(rx_image)
     preds = model.predict(img_array)
     score = tf.nn.softmax(preds[0])
@@ -29,16 +29,23 @@ def predict(rx_image, keras_model_file):
     other_class = CLASS_NAMES[other_index]
     other_confidence = score[other_index] * 100
     
-    return f"Resultado: {predicted_class.upper()} ({confidence:.2f}%)\n" + \
-           f"Probabilidad {other_class}: {other_confidence:.2f}%"
+    # Crear texto de salida detallado
+    resultado_texto = (
+        f"--- Resultado de la Clasificación ---\n"
+        f"La imagen es: **{predicted_class.upper()}**\n"
+        f"Confianza: {confidence:.2f}%\n"
+        f"Probabilidad de ser {other_class}: {other_confidence:.2f}%\n"
+        f"------------------------------------"
+    )
+    
+    return rx_image, resultado_texto
 
+# Interfaz Gradio
 iface = gr.Interface(
     fn=predict,
-    inputs=[
-        gr.Image(type="numpy", label="Sube tu RX"),
-        gr.File(file_types=[".keras"], label="Sube tu modelo Keras")
-    ],
-    outputs=gr.Textbox(label="Resultado")
+    inputs=gr.Image(type="numpy", label="Sube tu RX"),
+    outputs=[gr.Image(type="numpy", label="RX procesada"), gr.Textbox(label="Resultado")],
+    live=False
 )
 
 iface.launch()