Perceptual Grouping in Contrastive Vision-Language Models

Recent advances in zero-shot image recognition suggest that vision-language models learn generic visual representations with a high degree of semantic information that may be arbitrarily probed with natural language phrases. Understanding an image, however, is not just about understanding what content resides within an image, but importantly, where that content resides. In this work we examine how well vision-language models are able to understand where objects reside within an image and group together visually related parts of the imagery. We demonstrate how contemporary vision and language representation learning models based on contrastive losses and large web-based data capture limited object localization information. We propose a minimal set of modifications that results in models that uniquely learn both semantic and spatial information. We measure this performance in terms of zero-shot image recognition, unsupervised bottom-up and top-down semantic segmentations, as well as robustness analyses. We find that the resulting model achieves state-of-the-art results in terms of unsupervised segmentation, and demonstrate that the learned representations are uniquely robust to spurious correlations in datasets designed to probe the causal behavior of vision models.

PDF Abstract ICCV 2023 PDF ICCV 2023 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Unsupervised Semantic Segmentation with Language-image Pre-training ADE20K CLIPpy ViT-B Mean IoU (val) 13.5 # 6
Unsupervised Semantic Segmentation with Language-image Pre-training Cityscapes val CLIPpy ViT-B mIoU 18.1 # 9
Unsupervised Semantic Segmentation with Language-image Pre-training MS COCO CLIPpy ViT-B Mean IoU (val) 25.5 # 1
Unsupervised Semantic Segmentation with Language-image Pre-training PASCAL VOC 2007 CLIPpy ViT-B Mean IoU (val) 52.2 # 1

Methods


No methods listed for this paper. Add relevant methods here