Update app.py

app.py

@@ -2,45 +2,32 @@ import gradio as gr
 import tensorflow as tf
 import numpy as np
 
-# Define a custom layer for FixedDropout
-class FixedDropout(tf.keras.layers.Dropout):
-    def _get_noise_shape(self, inputs):
-        if self.noise_shape is None:
-            return self.noise_shape
-        symbolic_shape = tf.shape(inputs)
-        noise_shape = [symbolic_shape[axis] if shape is None else shape
-                       for axis, shape in enumerate(self.noise_shape)]
-        return tuple(noise_shape)
-
-# Register the custom layer
-tf.keras.utils.get_custom_objects()['FixedDropout'] = FixedDropout
-
 # Load your trained TensorFlow model
 model = tf.keras.models.load_model('modelo_treinado.h5')  # Load your saved model
 
 # Define a function to make predictions
 def classify_image(input_image):
-    # Preprocess the input image (resize and normalize)
-    input_image = tf.image.resize(input_image, (224, 224))  # Make sure to match your model's input size
-    input_image = (input_image / 255.0)  # Normalize to [0, 1]
-    input_image = np.expand_dims(input_image, axis=0)  # Add batch dimension
+    # Redimensione a imagem para as dimensões corretas (192x256)
+    input_image = tf.image.resize(input_image, (192, 256))  # Redimensione para as dimensões esperadas
+    input_image = (input_image / 255.0)  # Normalize para [0, 1]
+    input_image = np.expand_dims(input_image, axis=0)  # Adicione a dimensão de lote
 
-    # Make a prediction using your model
+    # Faça a previsão usando o modelo
     prediction = model.predict(input_image)
 
-    # Assuming your model outputs probabilities for two classes, you can return the class with the highest probability
+    # Assumindo que o modelo retorna probabilidades para duas classes, você pode retornar a classe com a maior probabilidade
     class_index = np.argmax(prediction)
-    class_labels = ["Normal", "Cataract"]  # Replace with your actual class labels
+    class_labels = ["Normal", "Cataract"]  # Substitua pelas suas etiquetas de classe reais
     predicted_class = class_labels[class_index]
 
     return predicted_class
 
-# Create a Gradio interface
+# Crie uma interface Gradio
 input_interface = gr.Interface(
     fn=classify_image,
-    inputs="image",  # Specify input type as "image"
-    outputs="text"   # Specify output type as "text"
+    inputs="image",  # Especifique o tipo de entrada como "image"
+    outputs="text"   # Especifique o tipo de saída como "text"
 )
 
-# Launch the Gradio app
+# Inicie o aplicativo Gradio
 input_interface.launch()