Saved in:
Bibliographic Details
Main Authors: Shao, Tong, Tian, Zhuotao, Zhao, Hang, Su, Jingyong
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2407.08268
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916320013975552
author Shao, Tong
Tian, Zhuotao
Zhao, Hang
Su, Jingyong
author_facet Shao, Tong
Tian, Zhuotao
Zhao, Hang
Su, Jingyong
contents CLIP, as a vision-language model, has significantly advanced Open-Vocabulary Semantic Segmentation (OVSS) with its zero-shot capabilities. Despite its success, its application to OVSS faces challenges due to its initial image-level alignment training, which affects its performance in tasks requiring detailed local context. Our study delves into the impact of CLIP's [CLS] token on patch feature correlations, revealing a dominance of "global" patches that hinders local feature discrimination. To overcome this, we propose CLIPtrase, a novel training-free semantic segmentation strategy that enhances local feature awareness through recalibrated self-correlation among patches. This approach demonstrates notable improvements in segmentation accuracy and the ability to maintain semantic coherence across objects.Experiments show that we are 22.3% ahead of CLIP on average on 9 segmentation benchmarks, outperforming existing state-of-the-art training-free methods.The code are made publicly available at: https://github.com/leaves162/CLIPtrase.
format Preprint
id arxiv_https___arxiv_org_abs_2407_08268
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Explore the Potential of CLIP for Training-Free Open Vocabulary Semantic Segmentation
Shao, Tong
Tian, Zhuotao
Zhao, Hang
Su, Jingyong
Computer Vision and Pattern Recognition
CLIP, as a vision-language model, has significantly advanced Open-Vocabulary Semantic Segmentation (OVSS) with its zero-shot capabilities. Despite its success, its application to OVSS faces challenges due to its initial image-level alignment training, which affects its performance in tasks requiring detailed local context. Our study delves into the impact of CLIP's [CLS] token on patch feature correlations, revealing a dominance of "global" patches that hinders local feature discrimination. To overcome this, we propose CLIPtrase, a novel training-free semantic segmentation strategy that enhances local feature awareness through recalibrated self-correlation among patches. This approach demonstrates notable improvements in segmentation accuracy and the ability to maintain semantic coherence across objects.Experiments show that we are 22.3% ahead of CLIP on average on 9 segmentation benchmarks, outperforming existing state-of-the-art training-free methods.The code are made publicly available at: https://github.com/leaves162/CLIPtrase.
title Explore the Potential of CLIP for Training-Free Open Vocabulary Semantic Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2407.08268