app.py

import gradio as gr
import torch
import torchvision.transforms as transforms
from PIL import Image
from resnet import SupCEResNet  # Ensure the correct import path

# ✅ Define class labels (from Clothing1M)
class_labels = [
    "T-Shirt", "Shirt", "Knitwear", "Chiffon", "Sweater", "Hoodie",
    "Windbreaker", "Jacket", "Downcoat", "Suit", "Shawl", "Dress",
    "Vest", "Underwear"
]

# ✅ Function to load the model
def create_model_selfsup(net='resnet50', num_class=14, checkpoint_path='/content/ckpt_clothing_resnet50.pth'):
    """Loads a self-supervised pretrained model for Clothing1M classification"""
    print(f"🔄 Loading model from: {checkpoint_path}")

    # Load the checkpoint safely
    checkpoint = torch.load(checkpoint_path, map_location="cuda" if torch.cuda.is_available() else "cpu", weights_only=False)

    # Remove 'module.' prefix if using DataParallel
    state_dict = {k.replace('module.', ''): v for k, v in checkpoint['model'].items()}

    # Initialize and load model
    model = SupCEResNet(net, num_classes=num_class, pool=True)
    model.load_state_dict(state_dict, strict=False)

    # Move model to GPU if available
    model = model.to("cuda" if torch.cuda.is_available() else "cpu")
    model.eval()  # Set model to evaluation mode

    print("✅ Model loaded successfully!")
    return model

# ✅ Load the model once
model = create_model_selfsup()

# ✅ Define image preprocessing function
def preprocess_image(image):
    """Transforms input image for the model"""
    transform = transforms.Compose([
        transforms.Resize(256),
        transforms.CenterCrop(224),
        transforms.ToTensor(),
        transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
    ])
    return transform(image).unsqueeze(0).to("cuda" if torch.cuda.is_available() else "cpu")

# ✅ Define inference function
def predict_clothing(image):
    """Runs inference on an uploaded image"""
    image = Image.fromarray(image)  # Convert numpy array to PIL Image
    image = preprocess_image(image)  # Preprocess image

    with torch.no_grad():
        output = model(image)
        predicted_class = torch.argmax(output, dim=1).item()  # Get class index

    return class_labels[predicted_class]  # Return class name

# ✅ Create Gradio Interface
gr.Interface(
    fn=predict_clothing,
    inputs=gr.Image(type="numpy"),
    outputs=gr.Textbox(label="Predicted Clothing Type"),
    title="Clothing1M Classification",
    description="Upload an image to classify clothing into one of 14 categories."
).launch()