Test-time Contrastive Concepts for Open-world Semantic Segmentation with Vision-Language Models

TL;DR

This paper introduces test-time contrastive concepts for open-world semantic segmentation using vision-language models, enabling segmentation of a single concept from a textual prompt without exhaustive concept lists.

Contribution

It proposes two methods to generate test-time contrastive concepts leveraging training set text distribution and LLM prompts, improving segmentation in realistic scenarios.

Findings

Effective segmentation of single concepts with minimal prior info

Test-time contrastive concepts outperform baseline methods

New evaluation metric for single-concept segmentation

Abstract

Recent CLIP-like Vision-Language Models (VLMs), pre-trained on large amounts of image-text pairs to align both modalities with a simple contrastive objective, have paved the way to open-vocabulary semantic segmentation. Given an arbitrary set of textual queries, image pixels are assigned the closest query in feature space. However, this works well when a user exhaustively lists all possible visual concepts in an image that contrast against each other for the assignment. This corresponds to the current evaluation setup in the literature, which relies on having access to a list of in-domain relevant concepts, typically classes of a benchmark dataset. Here, we consider the more challenging (and realistic) scenario of segmenting a single concept, given a textual prompt and nothing else. To achieve good results, besides contrasting with the generic 'background' text, we propose two different approaches to automatically generate, at test time, query-specific textual contrastive concepts. We do so by leveraging the distribution of text in the VLM's training set or crafted LLM prompts. We also propose a metric designed to evaluate this scenario and show the relevance of our approach on commonly used datasets.