TeDA: Boosting Vision-Lanuage Models for Zero-Shot 3D Object Retrieval via Testing-time Distribution Alignment

TL;DR

This paper introduces TeDA, a novel test-time adaptation framework that enhances zero-shot 3D object retrieval by aligning 3D features with pre-trained vision-language models like CLIP, incorporating multi-view projections and textual descriptions.

Contribution

TeDA is the first approach to adapt a pretrained vision-language model for 3D object retrieval at test time, significantly improving zero-shot performance without extensive training.

Findings

TeDA outperforms state-of-the-art methods on four 3D retrieval benchmarks.

Incorporating textual descriptions further boosts retrieval accuracy.

Effective even with limited training data and using depth maps.

Abstract

Learning discriminative 3D representations that generalize well to unknown testing categories is an emerging requirement for many real-world 3D applications. Existing well-established methods often struggle to attain this goal due to insufficient 3D training data from broader concepts. Meanwhile, pre-trained large vision-language models (e.g., CLIP) have shown remarkable zero-shot generalization capabilities. Yet, they are limited in extracting suitable 3D representations due to substantial gaps between their 2D training and 3D testing distributions. To address these challenges, we propose Testing-time Distribution Alignment (TeDA), a novel framework that adapts a pretrained 2D vision-language model CLIP for unknown 3D object retrieval at test time. To our knowledge, it is the first work that studies the test-time adaptation of a vision-language model for 3D feature learning. TeDA projects 3D objects into multi-view images, extracts features using CLIP, and refines 3D query embeddings with an iterative optimization strategy by confident query-target sample pairs in a self-boosting manner. Additionally, TeDA integrates textual descriptions generated by a multimodal language model (InternVL) to enhance 3D object understanding, leveraging CLIP's aligned feature space to fuse visual and textual cues. Extensive experiments on four open-set 3D object retrieval benchmarks demonstrate that TeDA greatly outperforms state-of-the-art methods, even those requiring extensive training. We also experimented with depth maps on Objaverse-LVIS, further validating its effectiveness. Code is available at https://github.com/wangzhichuan123/TeDA.