Paper ID: 2411.10086

CorrCLIP: Reconstructing Correlations in CLIP with Off-the-Shelf Foundation Models for Open-Vocabulary Semantic Segmentation

Dengke Zhang, Fagui Liu, Quan Tang

Open-vocabulary semantic segmentation aims to assign semantic labels to each pixel without relying on a predefined set of categories. Contrastive Language-Image Pre-training (CLIP) demonstrates outstanding zero-shot classification capabilities but struggles with the pixel-wise segmentation task as the captured inter-patch correlations correspond to no specific visual concepts. Despite previous CLIP-based works improving inter-patch correlations by self-self attention, they still face the inherent limitation that image patches tend to have high similarity to outlier ones. In this work, we introduce CorrCLIP, a training-free approach for open-vocabulary semantic segmentation, which reconstructs significantly coherent inter-patch correlations utilizing foundation models. Specifically, it employs the Segment Anything Model (SAM) to define the scope of patch interactions, ensuring that patches interact only with semantically similar ones. Furthermore, CorrCLIP obtains an understanding of an image's semantic layout via self-supervised models to determine concrete similarity values between image patches, which addresses the similarity irregularity problem caused by the aforementioned restricted patch interaction regime. Finally, CorrCLIP reuses the region masks produced by SAM to update the segmentation map. As a training-free method, CorrCLIP achieves a notable improvement across eight challenging benchmarks regarding the averaged mean Intersection over Union, boosting it from 44.4% to 51.0%.

Submitted: Nov 15, 2024