app.py

import gradio as gr
import math
import os
import cv2
import numpy as np
import torch
import segmentation_models_pytorch as smp

def pad_to_divisible(img, div=32):
    h, w, _ = img.shape
    new_h = math.ceil(h / div) * div
    new_w = math.ceil(w / div) * div
    pad_bottom = new_h - h
    pad_right = new_w - w
    padded = cv2.copyMakeBorder(img, 0, pad_bottom, 0, pad_right, cv2.BORDER_CONSTANT, value=[0, 0, 0])
    return padded


device = torch.device("cpu")
print("Using device:", device)


model_path = "best_unet.pth"
if os.path.exists(model_path):
    model = smp.Unet(
        encoder_name="resnet34",
        encoder_weights=None,
        in_channels=3,
        classes=1
    )
    model.load_state_dict(torch.load(model_path, map_location=device))
    model.to(device)
    model.eval()
    print("✅ Model loaded successfully.")
else:
    raise FileNotFoundError(f"❌ Model file not found at: {model_path}")

def predict(image):
    if image.shape[2] == 4:
        image = cv2.cvtColor(image, cv2.COLOR_RGBA2RGB)

    original = image.copy()
    padded = pad_to_divisible(original, div=32)

    normalized = padded.astype(np.float32) / 255.0
    tensor = torch.from_numpy(normalized).permute(2, 0, 1).unsqueeze(0).to(device)

    with torch.no_grad():
        output = model(tensor)
        mask = (torch.sigmoid(output) > 0.3).float().cpu().squeeze().numpy()

    overlay = padded.astype(np.float32)
    mask_bool = mask > 0.5
    red = np.array([255, 0, 0], dtype=np.float32)
    overlay[mask_bool] = (1 - 0.5) * overlay[mask_bool] + 0.5 * red
    overlay = np.clip(overlay, 0, 255).astype(np.uint8)

    return padded, mask, overlay

demo = gr.Interface(
    fn=predict,
    inputs=gr.Image(type="numpy", label="Upload Face Image"),
    outputs=[
        gr.Image(label="Padded Input"),
        gr.Image(label="Predicted Mask"),
        gr.Image(label="Overlay on Image")
    ],
    title="Wrinkle Segmentation",
    description="Upload a face image to see wrinkle regions detected"
)

demo.launch()