Enhancing Few-Shot Out-of-Distribution Detection via the Refinement of Foreground and Background

TL;DR

This paper introduces a plug-and-play framework that enhances few-shot out-of-distribution detection by adaptively suppressing background patches and rectifying confusable foreground patches, building on existing foreground-background decomposition methods.

Contribution

It proposes a novel framework with three modules to address limitations in current FG-BG methods, significantly improving OOD detection performance.

Findings

Significant performance improvements on OOD detection benchmarks.

Effective adaptive weighting of background patches improves detection accuracy.

Rectification of confusable foreground patches reduces misclassification.

Abstract

CLIP-based foreground-background (FG-BG) decomposition methods have demonstrated remarkable effectiveness in improving few-shot out-of-distribution (OOD) detection performance. However, existing approaches still suffer from several limitations. For background regions obtained from decomposition, existing methods adopt a uniform suppression strategy for all patches, overlooking the varying contributions of different patches to the prediction. For foreground regions, existing methods fail to adequately consider that some local patches may exhibit appearance or semantic similarity to other classes, which may mislead the training process. To address these issues, we propose a new plug-and-play framework. This framework consists of three core components: (1) a Foreground-Background Decomposition module, which follows previous FG-BG methods to separate an image into foreground and background regions; (2) an Adaptive Background Suppression module, which adaptively weights patch classification entropy; and (3) a Confusable Foreground Rectification module, which identifies and rectifies confusable foreground patches. Extensive experimental results demonstrate that the proposed plug-and-play framework significantly improves the performance of existing FG-BG decomposition methods. Code is available at: https://github.com/lounwb/FoBoR.