Everything Can Be Described in Words: A Simple Unified Multi-Modal Framework with Semantic and Temporal Alignment

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bi, Xiaowei, Xu, Zheyuan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908401646174208
author Bi, Xiaowei
Xu, Zheyuan
author_facet Bi, Xiaowei
Xu, Zheyuan
contents While multi-modal learning has advanced significantly, current approaches often create inconsistencies in representation and reasoning of different modalities. We propose UMaT, a theoretically-grounded framework that unifies visual and auditory inputs as structured text for large language models, addressing semantic alignment, temporal synchronization, and efficient sparse information retrieval. It significantly improves state-of-the-art Long Video Question Answering accuracy (up to 13.7%, and 16.9% on long videos) via redundancy minimization and structured textual representation for unified multi-modal reasoning
format Preprint
id arxiv_https___arxiv_org_abs_2503_09081
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Everything Can Be Described in Words: A Simple Unified Multi-Modal Framework with Semantic and Temporal Alignment
Bi, Xiaowei
Xu, Zheyuan
Computer Vision and Pattern Recognition
Artificial Intelligence
While multi-modal learning has advanced significantly, current approaches often create inconsistencies in representation and reasoning of different modalities. We propose UMaT, a theoretically-grounded framework that unifies visual and auditory inputs as structured text for large language models, addressing semantic alignment, temporal synchronization, and efficient sparse information retrieval. It significantly improves state-of-the-art Long Video Question Answering accuracy (up to 13.7%, and 16.9% on long videos) via redundancy minimization and structured textual representation for unified multi-modal reasoning
title Everything Can Be Described in Words: A Simple Unified Multi-Modal Framework with Semantic and Temporal Alignment
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2503.09081