CLIPSeg: Image Segmentation Using Text and Image Prompts

CLIPSeg: Image Segmentation Using Text and Image Prompts | HuggingFace docs

Demo for using CLIPSeg, a CLIP-based model for zero- and one-shot image segmentation. To use it, simply upload an image and add a text to mask (identify in the image), or use one of the examples below and click 'submit'. Results will show up in a few seconds.

Example images are taken from the ImageNet-A dataset

Image

Please describe what you want to identify

Threshold

0 1

Alpha

0 1

Draw rectangles

Segmentation Result

Mask

Output Image

Examples

	Please describe what you want to identify	Threshold	Alpha	Draw rectangles