English

X-VILA: Cross-Modality Alignment for Large Language Model

Computer Vision and Pattern Recognition 2024-05-30 v1 Computation and Language Machine Learning

Abstract

We introduce X-VILA, an omni-modality model designed to extend the capabilities of large language models (LLMs) by incorporating image, video, and audio modalities. By aligning modality-specific encoders with LLM inputs and diffusion decoders with LLM outputs, X-VILA achieves cross-modality understanding, reasoning, and generation. To facilitate this cross-modality alignment, we curate an effective interleaved any-to-any modality instruction-following dataset. Furthermore, we identify a significant problem with the current cross-modality alignment method, which results in visual information loss. To address the issue, we propose a visual alignment mechanism with a visual embedding highway module. We then introduce a resource-efficient recipe for training X-VILA, that exhibits proficiency in any-to-any modality conversation, surpassing previous approaches by large margins. X-VILA also showcases emergent properties across modalities even in the absence of similar training data. The project will be made open-source.

Keywords

Cite

@article{arxiv.2405.19335,
  title  = {X-VILA: Cross-Modality Alignment for Large Language Model},
  author = {Hanrong Ye and De-An Huang and Yao Lu and Zhiding Yu and Wei Ping and Andrew Tao and Jan Kautz and Song Han and Dan Xu and Pavlo Molchanov and Hongxu Yin},
  journal= {arXiv preprint arXiv:2405.19335},
  year   = {2024}
}

Comments

Technical Report

R2 v1 2026-06-28T16:46:05.980Z