Negative Label Guided OOD Detection with Pretrained Vision-Language Models
This addresses the problem of identifying unknown samples for trustworthy AI systems, offering a novel approach that leverages text modality in vision-language models, though it is incremental as it builds on existing OOD detection research.
The paper tackles out-of-distribution (OOD) detection in vision-language models by proposing NegLabel, a post hoc method that uses negative labels from corpus databases, achieving state-of-the-art performance on various benchmarks and demonstrating robustness against domain shifts.
Out-of-distribution (OOD) detection aims at identifying samples from unknown classes, playing a crucial role in trustworthy models against errors on unexpected inputs. Extensive research has been dedicated to exploring OOD detection in the vision modality. Vision-language models (VLMs) can leverage both textual and visual information for various multi-modal applications, whereas few OOD detection methods take into account information from the text modality. In this paper, we propose a novel post hoc OOD detection method, called NegLabel, which takes a vast number of negative labels from extensive corpus databases. We design a novel scheme for the OOD score collaborated with negative labels. Theoretical analysis helps to understand the mechanism of negative labels. Extensive experiments demonstrate that our method NegLabel achieves state-of-the-art performance on various OOD detection benchmarks and generalizes well on multiple VLM architectures. Furthermore, our method NegLabel exhibits remarkable robustness against diverse domain shifts. The codes are available at https://github.com/tmlr-group/NegLabel.