app.py

import spaces
import gradio as gr

from PIL import Image

import torch
from diffusion import DiffusionPipeline

device = "cuda" if torch.cuda.is_available() else "cpu"

pipe = DiffusionPipeline(device)

def read_content(file_path: str) -> str:
    """read the content of target file
    """
    with open(file_path, 'r', encoding='utf-8') as f:
        content = f.read()

    return content

@spaces.GPU
def predict(input, dkernel, diffusion_step, q=False):
    lq = input["image"].convert("RGB")
    mask = input["mask"].convert("RGB")
    mask = mask.resize(lq.size, resample=Image.NEAREST)
    output = pipe(lq=lq, mask=mask, dkernel=dkernel, diffusion_step=diffusion_step)
    return output

@spaces.GPU
def qpredict(input, dkernel, diffusion_step, q=False):
    lq = input["image"].convert("RGB")
    mask = input["mask"].convert("RGB")
    mask = mask.resize(lq.size, resample=Image.NEAREST)
    for output in pipe.quick_solve(lq=lq, mask=mask, dkernel=dkernel, diffusion_step=diffusion_step):
        yield output


css = '''
.container {max-width: 1150px;margin: auto;padding-top: 1.5rem}
#image_upload{min-height:400px}
#image_upload [data-testid="image"], #image_upload [data-testid="image"] > div{min-height: 400px}
#mask_radio .gr-form{background:transparent; border: none}
#word_mask{margin-top: .75em !important}
#word_mask textarea:disabled{opacity: 0.3}
.footer {margin-bottom: 45px;margin-top: 35px;text-align: center;border-bottom: 1px solid #e5e5e5}
.footer>p {font-size: .8rem; display: inline-block; padding: 0 10px;transform: translateY(10px);background: white}
.dark .footer {border-color: #303030}
.dark .footer>p {background: #0b0f19}
.acknowledgments h4{margin: 1.25em 0 .25em 0;font-weight: bold;font-size: 115%}
#image_upload .touch-none{display: flex}
@keyframes spin {
    from {
        transform: rotate(0deg);
    }
    to {
        transform: rotate(360deg);
    }
}
'''

image_blocks = gr.Blocks(css=css)
with image_blocks as demo:
    gr.HTML(read_content("header.html"))
    with gr.Group():
        with gr.Group():
            with gr.Row():
                with gr.Column():
                    im = gr.ImageEditor(
                        type="pil",
                        crop_size="1:1",
                        elem_id="image_upload",
                        label="Shadow Image"
                    ).style(height=400)
                    # image = gr.Image(source='upload', tool='sketch', elem_id="image_upload", type="pil", label="Shadow Image").style(height=400)
                    dkernel = gr.Slider(minimum=11, maximum=55, step=2, value=11, label="Dilation Kernel Size")
                    diffusion_step = gr.Slider(minimum=10, maximum=200, step=5, value=20, label="Diffusion Time Step")
                    with gr.Row(elem_id="prompt-container").style(mobile_collapse=False, equal_height=True):
                        with gr.Column():
                            btn = gr.Button("Removal").style(
                                margin=False,
                                full_width=True,
                            )
                        with gr.Column():
                            qbtn = gr.Button("Quick Removal").style(
                                margin=False,
                                full_width=True,
                            )

                with gr.Column():
                    image_out = gr.Image(label="Removal Result", elem_id="output-img")
            with gr.Row():
                gr.Examples(examples=[
                    'examples/lssd2025.jpg',
                    'examples/web-shadow0248.jpg',
                ], inputs=[image])

            btn.click(fn=predict, inputs=[image, dkernel, diffusion_step], outputs=[image_out])
            qbtn.click(fn=qpredict, inputs=[image, dkernel, diffusion_step], outputs=[image_out])

image_blocks.launch(enable_queue=True)