Paper Image

Text-guided referring image segmentation

Published on:

27 October 2023

Primary Category:

Computer Vision and Pattern Recognition

Paper Authors:

Yucheng Suo,

Linchao Zhu,

Yi Yang


Key Details

Proposes training-free referring image segmentation using CLIP

Generates text to enhance fine-grained region-text alignment

Uses positive and negative texts to improve matching

Outperforms previous state-of-the-art methods

Does not require model architecture changes

AI generated summary

Text-guided referring image segmentation

This paper proposes a new framework for referring image segmentation that does not require pixel-level supervision during training. It introduces techniques to enhance the alignment between referring expressions and local image regions using generated text, improving results.

Answers from this paper


No comments yet, be the first to start the conversation...

Sign up to comment on this paper

Sign Up