app.py

import gradio as gr
from PIL import Image
from torchvision import transforms

from explain import do_explain

normalize = transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5])
TRANSFORM = transforms.Compose(
    [
        transforms.Resize(256),
        transforms.CenterCrop(224),
        transforms.ToTensor(),
        normalize,
    ]
)


def generate_viz(image, class_index=None):
    if class_index is not None:
        class_index = int(class_index)
    print(f"Image: {image.size}")
    print(f"Class: {class_index}")
    
    viz, pred = do_explain(TRANSFORM, image, class_index=class_index)
    viz.savefig("visualization.png")
    return Image.open("visualization.png").convert("RGB"), pred

title = "Explain ViT 😊"

iface = gr.Interface(fn=generate_viz, inputs=[
        gr.Image(type="pil", label="Input Image"),
       
        gr.Number(label="Class Index", info="Class index to explain"),
    ],
    outputs=[ gr.Image(label="XAI-Image"), gr.Text(label="prob"),],
    title=title,
    allow_flagging="never",
    cache_examples=True,
    examples=[
        ["ViT_DeiT/samples/catdog.png",None],
        ["ViT_DeiT/samples/catdog.png", 243],
        ["ViT_DeiT/samples/el2.png", None],
        ["ViT_DeiT/samples/el2.png", 340],
    ],
)

iface.launch(debug=True)