MMOne: Representing Multiple Modalities in One Scene

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gu, Zhifeng, Wang, Bing
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916847481257984
author Gu, Zhifeng
Wang, Bing
author_facet Gu, Zhifeng
Wang, Bing
contents Humans perceive the world through multimodal cues to understand and interact with the environment. Learning a scene representation for multiple modalities enhances comprehension of the physical world. However, modality conflicts, arising from inherent distinctions among different modalities, present two critical challenges: property disparity and granularity disparity. To address these challenges, we propose a general framework, MMOne, to represent multiple modalities in one scene, which can be readily extended to additional modalities. Specifically, a modality modeling module with a novel modality indicator is proposed to capture the unique properties of each modality. Additionally, we design a multimodal decomposition mechanism to separate multi-modal Gaussians into single-modal Gaussians based on modality differences. We address the essential distinctions among modalities by disentangling multimodal information into shared and modality-specific components, resulting in a more compact and efficient multimodal scene representation. Extensive experiments demonstrate that our method consistently enhances the representation capability for each modality and is scalable to additional modalities. The code is available at https://github.com/Neal2020GitHub/MMOne.
format Preprint
id arxiv_https___arxiv_org_abs_2507_11129
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MMOne: Representing Multiple Modalities in One Scene
Gu, Zhifeng
Wang, Bing
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Humans perceive the world through multimodal cues to understand and interact with the environment. Learning a scene representation for multiple modalities enhances comprehension of the physical world. However, modality conflicts, arising from inherent distinctions among different modalities, present two critical challenges: property disparity and granularity disparity. To address these challenges, we propose a general framework, MMOne, to represent multiple modalities in one scene, which can be readily extended to additional modalities. Specifically, a modality modeling module with a novel modality indicator is proposed to capture the unique properties of each modality. Additionally, we design a multimodal decomposition mechanism to separate multi-modal Gaussians into single-modal Gaussians based on modality differences. We address the essential distinctions among modalities by disentangling multimodal information into shared and modality-specific components, resulting in a more compact and efficient multimodal scene representation. Extensive experiments demonstrate that our method consistently enhances the representation capability for each modality and is scalable to additional modalities. The code is available at https://github.com/Neal2020GitHub/MMOne.
title MMOne: Representing Multiple Modalities in One Scene
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2507.11129