app.py

import gradio as gr
import tensorflow as tf
import numpy as np
from PIL import Image

from tensorflow.keras.applications import MobileNetV2
from tensorflow.keras.layers import Dense, BatchNormalization, Dropout
from tensorflow.keras.models import Sequential

# =====================
# MODEL ARCHITECTURE
# =====================

base_model = MobileNetV2(
    weights=None,
    include_top=False,
    input_shape=(224, 224, 3),
    pooling="avg"
)

model = Sequential([
    base_model,
    BatchNormalization(),
    Dropout(0.5),
    Dense(256, activation="relu"),
    Dropout(0.3),
    Dense(7, activation="softmax")
])

# Load trained weights
model.load_weights("model.weights.h5")

# =====================
# CLASS NAMES
# =====================

class_names = [
    "broken_benches",
    "fallen_trees",
    "garbage",
    "leaky_pipes",
    "open_manhole",
    "potholes",
    "streetlight"
]

# =====================
# PREDICTION FUNCTION
# =====================

def predict_image(image):
    if image is None:
        return "No image uploaded", "0%"

    image = image.convert("RGB")
    image = image.resize((224, 224))

    img = np.array(image, dtype=np.float32) / 255.0
    img = np.expand_dims(img, axis=0)

    prediction = model.predict(img, verbose=0)

    predicted_class = class_names[np.argmax(prediction)]
    confidence = float(np.max(prediction))

    return predicted_class, f"{confidence:.2%}"

# =====================
# UI
# =====================

description = """
# 🏙️ Community Issue Classification System

This AI-powered system automatically identifies common civic infrastructure issues from images.

### Detectable Categories
- 🪑 Broken Benches
- 🌳 Fallen Trees
- 🗑️ Garbage
- 🚰 Leaky Pipes
- ⚠️ Open Manholes
- 🕳️ Potholes
- 💡 Streetlight Issues

### Model Information
- Model: MobileNetV2 Fine-Tuned Classifier
- Classes: 7
- Input Size: 224 × 224 RGB Images
- Developer: Pauras More

Upload an image below to classify a civic issue.
"""

demo = gr.Interface(
    fn=predict_image,
    inputs=gr.Image(type="pil", label="Upload Image"),
    outputs=[
        gr.Textbox(label="Predicted Class"),
        gr.Textbox(label="Confidence")
    ],
    title="Community Issue Classifier",
    description=description,
    flagging_mode="never"
)

demo.launch()