A Short Overview of Multi-Modal Wi-Fi Sensing
Zijian Zhao
TL;DR
Wi‑Fi sensing leverages CSI and RSSI to enable contactless sensing within ISAC, but robustness across environments remains a core challenge. This survey catalogs the past 24 months of multi‑modal Wi‑Fi sensing literature, organizing methods into two main paradigms: multi‑modal fused sensing (input and feature fusion, plus mixture approaches) and multi‑modal enhanced training (cross‑modal knowledge distillation and ground‑truth label generation from strong modalities). It surveys a wide range of techniques, including cross‑modal transformers, CLIP‑based modal alignment, and vision‑assisted labeling, highlighting both performance gains and persistent hurdles like data scarcity and cross‑domain generalization. The work emphasizes future directions such as cross‑domain generalization, scalable pre‑training, open data/code, and foundation‑model style approaches to enable practical deployment while preserving privacy, efficiency, and adaptability; mathematically, the Wi‑Fi sensing model follows the classic signal relation $Y=HX+N$ with $H(f,t)=H_s(f,t)+H_d(f,t)$ and, for RSSI, $RSSI(t)=P_t-L(t)+\eta(t)$, illustrating the richer but more resource‑demanding information captured by CSI compared to RSSI.
Abstract
Wi-Fi sensing has emerged as a significant technology in wireless sensing and Integrated Sensing and Communication (ISAC), offering benefits such as low cost, high penetration, and enhanced privacy. Currently, it is widely utilized in various applications, including action recognition, human localization, and crowd counting. However, Wi-Fi sensing also faces challenges, such as low robustness and difficulties in data collection. Recently, there has been an increasing focus on multi-modal Wi-Fi sensing, where other modalities can act as teachers, providing ground truth or robust features for Wi-Fi sensing models to learn from, or can be directly fused with Wi-Fi for enhanced sensing capabilities. Although these methods have demonstrated promising results and substantial value in practical applications, there is a lack of comprehensive surveys reviewing them. To address this gap, this paper reviews the multi-modal Wi-Fi sensing literature \textbf{from the past 24 months} and highlights the current limitations, challenges and future directions in this field.