Update app.py

app.py

@@ -103,45 +103,73 @@
 # # Launch the interface
 # iface.launch()
 import gradio as gr
-from transformers import pipeline
-import numpy as np
-from PIL import Image
 import tensorflow as tf
+from tensorflow.keras.preprocessing import image as keras_image
+from PIL import Image
+import numpy as np
+import logging
 
-model_path = r'./model1_kera.h5'
-
-# Function to preprocess the image and make predictions
-def classify_LC_image(img):
-    try:
-        # Preprocess the image (resize, normalize, etc.)
-        img = image.resize((224,224))
-        img_array = tf.keras.preprocessing.image.img_to_array(img)
-        img_array = tf.expand_dims(img_array, 0)
+# Set up logging
+logging.basicConfig(level=logging.DEBUG)
 
+# Initialize the model variable
+model = None
 
-        # Making predictions
-        model_h5 = tf.keras.models.load_model(model_path)
-        predictions = model_h5.predict(img_array)
+# Load the trained model
+try:
+    model_path = './model1_kera.h5'  # Replace with the actual path to your model file
+    logging.info(f"Loading model from: {model_path}")
+    model = tf.keras.models.load_model(model_path)
+    logging.info("Model loaded successfully.")
+except Exception as e:
+    logging.error(f"Error loading model: {e}")
 
-        # Getting the class with the highest probability
-        class_idx = np.argmax(predictions)
-        class_label = ['Colon Adenocarcinoma', 'Colon Benign Tissue', 'Lung Adenocarcinoma', 'Lung Benign Tissue', 'Lung Squamous Cell Carcinoma'][class_idx]
-        confidence = predictions[0][class_idx]
+# Define the class names
+classes = ['Colon Adenocarcinoma', 'Colon Benign Tissue', 'Lung Adenocarcinoma', 'Lung Benign Tissue', 'Lung Squamous Cell Carcinoma']
 
-        return f"Prediction: {class_label}, Confidence: {confidence:.2f}"
+# Function to preprocess the uploaded image and make predictions
+def predict(img):
+    global model
+    try:
+        logging.debug("Received image for prediction.")
+        
+        # Resize and preprocess the image
+        img = img.resize((224, 224))
+        img_array = keras_image.img_to_array(img)
+        img_array = tf.expand_dims(img_array, 0)  # Add batch dimension
+        
+        logging.debug("Image preprocessed successfully.")
+        
+        # Ensure the model is loaded
+        if model is None:
+            raise ValueError("Model is not loaded properly.")
+        
+        # Make predictions
+        predictions = model.predict(img_array)
+        score = tf.nn.softmax(predictions[0])
+        predicted_class_index = tf.argmax(score).numpy()
+        predicted_class = classes[predicted_class_index]
+        
+        logging.debug(f"Prediction successful: {predicted_class}")
+        
+        # Return the predicted class and the raw predictions
+        return predicted_class, predictions[0].tolist()
     except Exception as e:
-        return str(e)
+        logging.error(f"Error during prediction: {e}")
+        # Print the error message in the output
+        return str(e), str(e)
 
-# Creating a Gradio interface
+# Create a Gradio interface
 iface = gr.Interface(
-    fn=classify_LC_image,
-    inputs="image",
-    outputs="text",
-    title="Alzheimer's Disease Classification",
-    description=" Upload an MRI Image for classification.",
-    # flagging_options = ["Wrong Prediction"],
-    # theme = 'darkhuggingface'
+    fn=predict, 
+    inputs=gr.Image(type='pil'), 
+    outputs=[
+        gr.Textbox(label="Prediction"),
+        gr.Label(label="Raw Predictions")
+    ], 
+    title="Lung and Colon Cancer Detection",
+    description="Upload an image of histopathological tissue to detect if it is a type of lung or colon cancer."
 )
 
-# Launching the Gradio interface
-iface.launch(inline = False)
+# Launch the interface
+iface.launch()