WiGRUNT: WiFi-Enabled Gesture Recognition Using Dual-Attention Network

Abstract

Gestures constitute an important form of nonverbal communication where bodily actions are used for delivering messages alone or in parallel with spoken words. Recently, there exists an emerging trend of WiFi sensing enabled gesture recognition due to its inherent merits like device-free, non-line-of-sight covering, and privacy-friendly. However, current WiFi-based approaches mainly reply on domain-specific training since they don't know "where to look" and "when to look ". To this end, we propose WiGRUNT, a WiFi-enabled gesture recognition system using dual-attention network, to mimic how a keen human being intercepting a gesture regardless of the environment variations. The key insight is to train the network to dynamically focus on the domain-independent features of a gesture on the WiFi Channel State Information (CSI) via a spatial-temporal dual-attention mechanism. WiGRUNT roots in a Deep Residual Network (ResNet) backbone to evaluate the importance of spatial-temporal clues and exploit their inbuilt sequential correlations for fine-grained gesture recognition. We evaluate WiGRUNT on the open Widar3 dataset and show that it significantly outperforms its state-of-the-art rivals by achieving the best-ever performance in-domain or cross-domain.

SenseFi: A library and benchmark on deep-learning-empowered WiFi human sensing

“… 59 2021 human activity recognition CNN Intel 5300 NIC few-shot learning Widar 37 2021 human identification, gesture recognition CNN-GRU Intel 5300 NIC supervised learning WiONE 60 2021 human identification CNN Intel 5300 NIC few-shot learning Ma et al. 61 2021 human activity recognition CNN, RNN, LSTM Intel 5300 NIC supervised learning THAT 62 2021 human activity recognition Transformers Intel 5300 NIC supervised learning WiGr 63 2021 gesture recognition CNN-LSTM Intel 5300 NIC supervised learning MCBAR 64 2021 human activity recognition CNN, GAN Atheros CSI Tool semi-supervised learning CAUTION 12 2022 human identification CNN Atheros CSI Tool few-shot learning CTS-AM 65 2022 human activity recognition CNN (attention) Intel 5300 NIC supervised learning WiGRUNT 66 2022 gesture recognition CNN (attention) Intel 5300 NIC supervised learning Zhuravchak et al. 67 2022 human activity recognition LSTM Nexmon CSI Tool supervised learning EfficientFi 39 …”

WiFi gesture recognition method based on data enhancement

“…The experimental results show that the recognition accuracy of cross-position is relatively stable, but in cross-direction gesture recognition, the prediction accuracy of different directions varies greatly, which also appears in WiDar3 and WiGRUNT [32] . The reason is that in cross-direction gesture recognition, gesture signals in direction 1 and direction 3 are blocked by human body.…”