LiteEmbed: Adapting CLIP to Rare Classes

TL;DR

LiteEmbed is a lightweight method that adapts CLIP to recognize rare or unseen classes by optimizing text embeddings without retraining the entire model, improving performance across various vision tasks.

Contribution

It introduces a PCA-based subspace-guided optimization for CLIP's text embeddings, enabling effective few-shot personalization without retraining encoders.

Findings

Significant performance improvements over prior methods.

Effective across classification, retrieval, segmentation, and detection tasks.

Seamless plug-and-play integration with CLIP.

Abstract

Large-scale vision-language models such as CLIP achieve strong zero-shot recognition but struggle with classes that are rarely seen during pretraining, including newly emerging entities and culturally specific categories. We introduce LiteEmbed, a lightweight framework for few-shot personalization of CLIP that enables new classes to be added without retraining its encoders. LiteEmbed performs subspace-guided optimization of text embeddings within CLIP's vocabulary, leveraging a PCA-based decomposition that disentangles coarse semantic directions from fine-grained variations. Two complementary objectives, coarse alignment and fine separation, jointly preserve global semantic consistency while enhancing discriminability among visually similar classes. Once optimized, the embeddings are plug-and-play, seamlessly substituting CLIP's original text features across classification, retrieval, segmentation, and detection tasks. Extensive experiments demonstrate substantial gains over prior methods, establishing LiteEmbed as an effective approach for adapting CLIP to underrepresented, rare, or unseen classes.