CodeBind: Decoupled Representation Learning for Multimodal Alignment with Unified Compositional Codebook

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Zeyu, Li, Jie, Han, Kai
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909054133075968
author Chen, Zeyu
Li, Jie
Han, Kai
author_facet Chen, Zeyu
Li, Jie
Han, Kai
contents Multimodal representation alignment is pivotal for large language models and robotics. Traditional methods are often hindered by cross-modal information discrepancies and data scarcity, leading to suboptimal alignment spaces that overlook modality-unique features. We propose CodeBind, a framework that optimizes multimodal representation spaces through a modality-shared-specific codebook design. By incrementally aligning target and bridging modalities, CodeBind bypasses the need for fully paired data. Unlike traditional hard alignment, CodeBind decomposes features into shared components for semantic consistency and specific components for modality-unique details. This design utilizes a compositional vector quantization scheme, where a shared codebook bridges modality gaps and modality-specific codebooks mitigate representation bias by preventing dominant modalities from overshadowing others. Validated across nine modalities (text, image, video, audio, depth, thermal, tactile, 3D point cloud, EEG), CodeBind achieves state-of-the-art performance in multimodal classification and retrieval tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2605_18257
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle CodeBind: Decoupled Representation Learning for Multimodal Alignment with Unified Compositional Codebook
Chen, Zeyu
Li, Jie
Han, Kai
Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
Multimodal representation alignment is pivotal for large language models and robotics. Traditional methods are often hindered by cross-modal information discrepancies and data scarcity, leading to suboptimal alignment spaces that overlook modality-unique features. We propose CodeBind, a framework that optimizes multimodal representation spaces through a modality-shared-specific codebook design. By incrementally aligning target and bridging modalities, CodeBind bypasses the need for fully paired data. Unlike traditional hard alignment, CodeBind decomposes features into shared components for semantic consistency and specific components for modality-unique details. This design utilizes a compositional vector quantization scheme, where a shared codebook bridges modality gaps and modality-specific codebooks mitigate representation bias by preventing dominant modalities from overshadowing others. Validated across nine modalities (text, image, video, audio, depth, thermal, tactile, 3D point cloud, EEG), CodeBind achieves state-of-the-art performance in multimodal classification and retrieval tasks.
title CodeBind: Decoupled Representation Learning for Multimodal Alignment with Unified Compositional Codebook
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2605.18257