File size: 870 Bytes
2e88d9b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 | import gradio as gr
from datasets import load_dataset
import numpy as np
# Load your dataset only once
ds = load_dataset("Devenarya/Microsoft100", split="train")
# Generator as before
def get_image_and_mask(index):
# This assumes dataset alternates: image, mask, image, mask, ...
images = [item['image'] for item in ds]
img = images[2 * index].convert('RGB')
mask = images[2 * index + 1].convert('L')
return np.array(img), np.array(mask)
def demo_fn(index):
# Output two images: the photo and its segmentation
return get_image_and_mask(index)
demo = gr.Interface(
fn=demo_fn,
inputs=gr.Slider(0, (len(ds)//2)-1, step=1, label="Image Index"),
outputs=[gr.Image(label="Original Image"), gr.Image(label="Segmentation Mask")],
title="Microsoft100 Image & Segmentation Explorer"
)
if __name__ == "__main__":
demo.launch()
|