CSI-IANet: An Inception Attention Network for Human-Human Interaction Recognition Based on CSI Signal

Abstract

In recent years, Wi-Fi infrastructures have become ubiquitous, providing device-free passivesensing features. Wi-Fi signals can be affected by their reflection, refraction, and absorption by moving objects in their path. The channel state information (CSI), a signal property indicator, of the Wi-Fi signal can be analyzed for human activity recognition (HAR). Deep learning-based HAR models can enhance performance and accuracy without sacrificing computational efficiency. However, to save computational power, an inception network, which uses a variety of techniques to boost speed and accuracy, can be adopted. In contrast, the concept of spatial attention can be applied to obtain refined features. In this paper, we propose a human-human interaction (HHI) classifier, CSI-IANet, which uses a modified inception CNN with a spatial-attention mechanism. The CSI-IANet consists of three steps: i) data processing, ii) feature extraction, and iii) recognition. The data processing layer first uses the second-order Butterworth low-pass filter to denoise the CSI signal and then segment it before feeding it to the model. The feature extraction layer uses a multilayer modified inception CNN with an attention mechanism that uses spatial attention in an intense structure to extract features from captured CSI signals. Finally, the refined features are exploited by the recognition section to determine HHIs correctly. To validate the performance of the proposed CSI-IANet, a publicly available HHI CSI dataset with a total of 4800 trials of 12 interactions was used. The performance of the proposed model was compared to those of existing state-of-the-art methods. The experimental results show that CSI-IANet achieved an average accuracy of 91.30%, which is better than that of the existing best method by 5%.

CSI-DeepNet: A Lightweight Deep Convolutional Neural Network Based Hand Gesture Recognition System Using Wi-Fi CSI Signal

“…The proposed model is trained and evaluated using a 10fold CV technique. The fold-wise performance results in Table 3 assumes that the highest accuracy is achieved for the To make a performance comparison with other deep learning based approaches, we use four different models: two pre-trained CNNs (ResNet-50 [29], DenseNet-121 [30]), an end-to-end deep learning framework (E2EDLF) [31], and CSI-IANet [32]. Two pre-trained CNNs are tuned via the transfer learning concept using the collected CSI gesture dataset.…”

CSI-Based Human Activity Recognition Using Multi-Input Multi-Output Autoencoder and Fine-Tuning

“…However, the presence of unsynchronized transmitters and receivers can cause random phase offsets in CSI and change it chaotically. In addition, the phase can be influenced by the sampling frequency offset, while CSI usually has an almost fixed range [ 24 ]. Therefore, the CSI amplitude is usually used.…”

A Lightweight Mobile Temporal Convolution Network for Multi-Location Human Activity Recognition based on Wi-Fi

“…With respect to deep learning-based methods, Kabir et al [15] introduced CSI-IANet, which utilized a Butterworth lowpass filter to denoise the CSI signal, employed three layers of CNN which is an inception module providing the model with varying receptive fields, and spatial-attention to first achieved an accuracy over 90% (91.3%). Kabir and Shin [28] presented DCNN, which employed only three layers of CNNs and achieved an accuracy of 88.66%.…”

“…Kabir et al [15] developed the CSI-based Inception Attention Network (CSI-IANet) incorporating CNNs and spatial-attention and evaluated it using a dataset of Wi-Fibased human-to-human interactions (HHI), which is the same dataset used in this paper [1]. The HHI dataset includes 12 different human-to-human interactions performed by two subjects and will be described in detail in section III.…”