VT-FSC: Vision-Tactile Fusion Semantic Communication for Remote Embodied Intelligence
Rui Xu, Yuanhang He, Gaolei Li, Chaofeng Zhang, Jianhua Li
Shanghai Jiao Tong University Advanced Institute of Industrial Technology
阅读操作
确认中在文库中上传 PDF 后可生成中文音频讲解。
摘要与影响
Remote embodied intelligence relies on the seamless transmission of rich multi-modal sensory information, particularly vision and touch, to enable intelligent agents to collaboratively perceive, interact with, and manipulate physical environments. However, transmitting such high-dimensional and heterogeneous data over wireless channels in real time poses substantial challenges in terms of bandwidth, latency, and semantic integrity. In this paper, we propose a novel vision-tactile fusion semantic communication (VT-FSC) framework tailored for remote embodied intelligence applications. By leveraging cross-modal feature fusion and attention-guided semantic compression, the proposed system transforms raw visual and tactile data into a unified low-dimensional semantic representation. This compact representation is then robustly transmitted through noisy wireless channels and decoded at the receiver to reconstruct both the visual scene and the tactile signal accurately. To ensure perceptual alignment, encoder and decoder are jointly optimized via human-in-the-loop feedback mechanisms. Experimental results in the multi-model dataset show that our method achieves up to 15% higher ST-SIM and 47% lower RMSE than baselines, validating the effectiveness of our framework in achieving high semantic compression rates and accurate perceptual reconstruction for remote embodied intelligence applications.
逐年被引趋势
关键指标
同类平均 = 1
同领域 · 同年份 · 同类型
Google Scholar 与 OpenAlex 的被引统计范围不同,数值存在差异属正常。
AI 辅助阅读
依据:摘要
可就本文提问;依据不足时会说明。
学术脉络
学科主题
工程Advanced Wireless Communication Technologies
Teleoperation and Haptic Systems · Video Coding and Compression Technologies