Towards Dexterous Embodied Manipulation via Deep Multi-Sensory Fusion and Sparse Expert Scaling

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Sun, Yirui, Zhuge, Guangyu, Liu, Keliang, Gu, Jie, xia, Zhihao, Ren, Qionglin, tian, Chunxu, Ga, Zhongxue
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910030110916608
author Sun, Yirui
Zhuge, Guangyu
Liu, Keliang
Gu, Jie
xia, Zhihao
Ren, Qionglin
tian, Chunxu
Ga, Zhongxue
author_facet Sun, Yirui
Zhuge, Guangyu
Liu, Keliang
Gu, Jie
xia, Zhihao
Ren, Qionglin
tian, Chunxu
Ga, Zhongxue
contents Realizing dexterous embodied manipulation necessitates the deep integration of heterogeneous multimodal sensory inputs. However, current vision-centric paradigms often overlook the critical force and geometric feedback essential for complex tasks. This paper presents DeMUSE, a Deep Multimodal Unified Sparse Experts framework leveraging a Diffusion Transformer to integrate RGB, depth, and 6-axis force into a unified serialized stream. Adaptive Modality-specific Normalization (AdaMN) is employed to recalibrate modality-aware features, mitigating representation imbalance and harmonizing the heterogeneous distributions of multi-sensory signals. To facilitate efficient scaling, the architecture utilizes a Sparse Mixture-of-Experts (MoE) with shared experts, increasing model capacity for physical priors while maintaining the low inference latency required for real-time control. A Joint denoising objective synchronously synthesizes environmental evolution and action sequences to ensure physical consistency. Achieving success rates of 83.2% and 72.5% in simulation and real-world trials, DeMUSE demonstrates state-of-the-art performance, validating the necessity of deep multi-sensory integration for complex physical interactions.
format Preprint
id arxiv_https___arxiv_org_abs_2602_19764
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Towards Dexterous Embodied Manipulation via Deep Multi-Sensory Fusion and Sparse Expert Scaling
Sun, Yirui
Zhuge, Guangyu
Liu, Keliang
Gu, Jie
xia, Zhihao
Ren, Qionglin
tian, Chunxu
Ga, Zhongxue
Robotics
Realizing dexterous embodied manipulation necessitates the deep integration of heterogeneous multimodal sensory inputs. However, current vision-centric paradigms often overlook the critical force and geometric feedback essential for complex tasks. This paper presents DeMUSE, a Deep Multimodal Unified Sparse Experts framework leveraging a Diffusion Transformer to integrate RGB, depth, and 6-axis force into a unified serialized stream. Adaptive Modality-specific Normalization (AdaMN) is employed to recalibrate modality-aware features, mitigating representation imbalance and harmonizing the heterogeneous distributions of multi-sensory signals. To facilitate efficient scaling, the architecture utilizes a Sparse Mixture-of-Experts (MoE) with shared experts, increasing model capacity for physical priors while maintaining the low inference latency required for real-time control. A Joint denoising objective synchronously synthesizes environmental evolution and action sequences to ensure physical consistency. Achieving success rates of 83.2% and 72.5% in simulation and real-world trials, DeMUSE demonstrates state-of-the-art performance, validating the necessity of deep multi-sensory integration for complex physical interactions.
title Towards Dexterous Embodied Manipulation via Deep Multi-Sensory Fusion and Sparse Expert Scaling
topic Robotics
url https://arxiv.org/abs/2602.19764