Device-Free Human Activity Recognition Based on GMM-HMM Using Channel State Information

Cheng, Xiaoyan; Huang, Binke; Zong, Jing

doi:10.1109/access.2021.3082627

Cited by 23 publications

(6 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To put it another way, in OFDM modulation, a single information stream is split among several closely spaced narrowband subchannel frequencies instead of a single wideband channel frequency. The presence of obstacles causes reflection, scattering, and multipath fading [ 20 ]. Therefore, when a person between the transmitter and receiver performs an activity, some changes will be made in transmitting the multipath of the Wi-Fi network.…”

Section: System Methodsmentioning

confidence: 99%

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

Chahoushi

Nabati

Asvadi

et al. 2023

Sensors

View full text Add to dashboard Cite

Wi-Fi-based human activity recognition (HAR) has gained considerable attention recently due to its ease of use and the availability of its infrastructures and sensors. Channel state information (CSI) captures how Wi-Fi signals are transmitted through the environment. Using channel state information of the received signals transmitted from Wi-Fi access points, human activity can be recognized with more accuracy compared with the received signal strength indicator (RSSI). However, in many scenarios and applications, there is a serious limit in the volume of training data because of cost, time, or resource constraints. In this study, multiple deep learning models have been trained for HAR to achieve an acceptable accuracy level while using less training data compared to other machine learning techniques. To do so, a pretrained encoder which is trained using only a limited number of data samples, is utilized for feature extraction. Then, by using fine-tuning, this encoder is utilized in the classifier, which is trained by a fraction of the rest of the data, and the training is continued alongside the rest of the classifier’s layers. Simulation results show that by using only 50% of the training data, there is a 20% improvement compared with the case where the encoder is not used. We also showed that by using an untrainable encoder, an accuracy improvement of 11% using 50% of the training data is achievable with a lower complexity level.

show abstract

Section: System Methodsmentioning

confidence: 99%

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

Chahoushi

Nabati

Asvadi

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

“…Chen et al [3], Shalaby et al [2], Li et al [21], and Muaaz et al [22] all employ 1,000 Hz sampling frequency for human activities successfully with varying techniques, number of participants, and number of activities. Cheng et al [23] used 800 Hz for human activity recognition successfully. Yang et al [24], Cui et al [25], Li et al [26], and Hwang et al [27] set the sampling rate to 500 Hz.…”

Section: B Sampling Rates In Recent Literaturementioning

confidence: 99%

Channel State Information for Human Activity Recognition with Low Sampling Rates

Brinke

Chiumento

Havinga

2023

2023 IEEE International Conference on Pervasive Computing and Communications Workshops and Other Affiliated Events (PerCom Work

View full text Add to dashboard Cite

In this paper, it is shown that lower transmission/sampling rates can be used in human activity recognition using channel state information (F1-scores > 85%) and that extremely high sampling rates are unnecessary once the system has been deployed. This is done by analysing the effects of interpolating different sampling rates on Wi-Fi dynamic channel state information for human activity recognition. While current research focuses on training and testing with homogeneous and very high sampling rates (> 100 Hz), this paper outlines some issues with higher sampling rates and explores the impact of training and testing with heterogeneous sampling rates in order to advance more towards joint communication and sensing, where one cannot be certain of the received data rate over time while not knowing the exact training set due to weight sharing in Federated Learning. This paper shows the effect of training and testing with heterogeneous sampling rates (including interpolated datasets) on convolutional neural networks in WiFi sensing.

show abstract

“…The insensitivity of wireless signals to the propagation medium and the widespread deployment of Wi-Fi devices have led to an increasing interest in behavior recognition techniques based on Wi-Fi devices. Behavior recognition techniques are based on analyzing the perturbation changes of human behavior on the wireless channel, using these changes to extract behavioral features and then constructing a behavioral classifier thus realizing human behavior recognition [1][2][3][4].…”

Section: Motivationmentioning

confidence: 99%

CSI-based Human Behavior Segmentation and Recognition Using Commodity Wi-Fi

Yang

Cheng

Tang

et al. 2022

Preprint

View full text Add to dashboard Cite

In recent years, the behavior recognition technology based on Wi-Fi devices has been favored by many researchers. Existing Wi-Fi-based human behavior recognition technology mainly uses classifification algorithms to construct classifification models,which has problems such as inaccurate behavior segmentation, failure to extract deep-level features from the original data and design classifification models matching the proposed features in the process of behavior recognition. In order to solve above problems, this paper proposes a window variance comparison method by combining the adaptive thresholds calculated to achieve effffective segmentation of multiple discontinuous human behaviors; then uses the short-time Fourier algorithm to extract time-frequency features of individual behavior, and extracts the graph structure data from the autocorrelation matrix of time-frequency features and the features themselves. A graph neural network is built for behavior recognition. The experimental results show that the segmentation accuracy of the behavior segmentation method in the two scenes is 0.964 and 0.993, which is better than the existing threshold-based behavior segmentation methods. In addition, this paper extracts graph structure data by spectral energy change method and builds behavior recognition model by using graph neural network, and the recognition accuracy is signifificantly improved compared with the traditional classifification algorithm.

show abstract

Device-Free Human Activity Recognition Based on GMM-HMM Using Channel State Information

Cited by 23 publications

References 21 publications

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

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

Channel State Information for Human Activity Recognition with Low Sampling Rates

CSI-based Human Behavior Segmentation and Recognition Using Commodity Wi-Fi

Contact Info

Product

Resources

About