Upload app.py with huggingface_hub

app.py

@@ -2,30 +2,38 @@ import gradio as gr
 import tensorflow as tf
 import numpy as np
 
-# Load model
-model = tf.keras.models.load_model("model.h5")
+# Image size used in training
+IMG_SIZE = (224, 224)
+
+# Final label mapping
+index_to_label = {
+    0: "paper",
+    1: "plastic",
+    2: "batteries",
+    3: "metal",
+    4: "glass",
+    5: "clothes",
+    6: "organic",
+    7: "light bulbs",
+    8: "e-waste"
+}
 
-# Labels
-class_names = ['batteries', 'clothes', 'e-waste', 'glass', 'light blubs',
-               'metal', 'organic', 'paper', 'plastic']
+# Load trained model
+model = tf.keras.models.load_model("model.h5")
 
-# Image preprocessing and prediction
-def predict_image(img):
-    img = tf.image.resize(img, (224, 224))
+# Image preprocessing function
+def classify_image(img):
+    img = tf.image.resize(img, IMG_SIZE)
+    img = tf.expand_dims(img, 0)
     img = tf.cast(img, tf.float32) / 255.0
-    img = tf.expand_dims(img, axis=0)
-    preds = model.predict(img)
-    pred_idx = np.argmax(preds)
-    confidence = float(np.max(preds)) * 100
-    return {class_names[pred_idx]: confidence}
+    preds = model.predict(img)[0]
+    return {index_to_label[i]: float(preds[i]) for i in range(len(preds))}
 
-# Gradio app
-demo = gr.Interface(
-    fn=predict_image,
-    inputs=gr.Image(type="numpy"),
+# Gradio Interface
+gr.Interface(
+    fn=classify_image,
+    inputs=gr.Image(type="numpy", label="Upload a Waste Image"),
     outputs=gr.Label(num_top_classes=3),
-    title="♻️ Waste Classifier",
-    description="Classify waste into 9 categories: batteries, clothes, e-waste, glass, light bulbs, metal, organic, paper, plastic."
-)
-
-demo.launch()
+    title="♻️ Smart Waste Classifier",
+    description="Upload an image of waste to classify it into 9 categories (paper, plastic, batteries, metal, glass, clothes, organic, light bulbs, e-waste). Built with TensorFlow + MobileNetV2"
+).launch()