Update app.py

app.py

@@ -1,29 +1,38 @@
 import gradio as gr
 import tensorflow as tf
 import numpy as np
+from PIL import Image
 
-# Load model with correct name
-model = tf.keras.models.load_model("waste_model_savedmodel")
-IMG_SIZE = (224, 224)
+# Load the model
+model = tf.keras.models.load_model("model.h5")
 
-# Class labels
-index_to_label = {
-    0: "paper", 1: "e-waste", 2: "metal", 3: "light blubs",
-    4: "organic", 5: "plastic", 6: "clothes", 7: "glass", 8: "batteries"
-}
+# Define the class labels (update based on your project)
+class_names = ['batteries', 'clothes', 'e-waste', 'glass', 'light blubs', 'metal', 'organic', 'paper', 'plastic']  # Example
 
-def classify_image(img):
-    img = tf.image.resize(img, IMG_SIZE)
-    img = tf.expand_dims(img, axis=0)
-    img = tf.cast(img, tf.float32) / 255.0
-    pred = model.predict(img)[0]
-    return {index_to_label[i]: float(pred[i]) for i in range(len(pred))}
+# Preprocess image
+def preprocess_image(image):
+    image = image.resize((224, 224))  # Adjust based on your model input
+    image = np.array(image) / 255.0   # Normalize if required
+    image = np.expand_dims(image, axis=0)  # Add batch dimension
+    return image
 
-demo = gr.Interface(
+# Prediction function
+def classify_image(image):
+    image = preprocess_image(image)
+    prediction = model.predict(image)
+    class_index = np.argmax(prediction)
+    class_name = class_names[class_index]
+    confidence = prediction[0][class_index]
+    return f"{class_name} ({confidence*100:.2f}%)"
+
+# Gradio interface
+interface = gr.Interface(
     fn=classify_image,
-    inputs=gr.Image(type="numpy", label="Upload Waste Image"),
-    outputs=gr.Label(num_top_classes=3),
-    title="Smart Waste Classifier"
+    inputs=gr.Image(type="pil"),
+    outputs="text",
+    title="Waste Image Classifier",
+    description="Upload an image of a recyclable item to classify it."
 )
 
-demo.launch()
+if __name__ == "__main__":
+    interface.launch()