Explore the Potential of CLIP for Training-Free Open Vocabulary Semantic Segmentation

TL;DR

This paper introduces CLIPtrase, a training-free method leveraging CLIP's features for improved open-vocabulary semantic segmentation, significantly enhancing accuracy without additional training.

Contribution

It proposes a novel, training-free approach that recalibrates patch correlations in CLIP to improve local feature discrimination for semantic segmentation.

Findings

Achieves 22.3% higher accuracy than CLIP on average across 9 benchmarks.

Outperforms existing state-of-the-art training-free segmentation methods.

Enhances local feature awareness and semantic coherence in segmentation tasks.

Abstract

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.