Papers
arxiv:2503.18817

Enhanced OoD Detection through Cross-Modal Alignment of Multi-Modal Representations

Published on Mar 24
· Submitted by mawjdgus on Apr 3

Abstract

Prior research on out-of-distribution detection (OoDD) has primarily focused on single-modality models. Recently, with the advent of large-scale pretrained vision-language models such as CLIP, OoDD methods utilizing such multi-modal representations through zero-shot and prompt learning strategies have emerged. However, these methods typically involve either freezing the pretrained weights or only partially tuning them, which can be suboptimal for downstream datasets. In this paper, we highlight that multi-modal fine-tuning (MMFT) can achieve notable OoDD performance. Despite some recent works demonstrating the impact of fine-tuning methods for OoDD, there remains significant potential for performance improvement. We investigate the limitation of na\"ive fine-tuning methods, examining why they fail to fully leverage the pretrained knowledge. Our empirical analysis suggests that this issue could stem from the modality gap within in-distribution (ID) embeddings. To address this, we propose a training objective that enhances cross-modal alignment by regularizing the distances between image and text embeddings of ID data. This adjustment helps in better utilizing pretrained textual information by aligning similar semantics from different modalities (i.e., text and image) more closely in the hyperspherical representation space. We theoretically demonstrate that the proposed regularization corresponds to the maximum likelihood estimation of an energy-based model on a hypersphere. Utilizing ImageNet-1k OoD benchmark datasets, we show that our method, combined with post-hoc OoDD approaches leveraging pretrained knowledge (e.g., NegLabel), significantly outperforms existing methods, achieving state-of-the-art OoDD performance and leading ID accuracy.

Community

Paper author Paper submitter

We propose a multi-modal fine-tuning (MMFT) approach to improve out-of-distribution detection (OoDD) in vision-language models. By addressing the modality gap between image and text embeddings through cross-modal alignment, our method enhances the utilization of pretrained knowledge, achieving state-of-the-art OoDD performance on ImageNet-1k benchmarks.

Your need to confirm your account before you can post a new comment.

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2503.18817 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2503.18817 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2503.18817 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.