2022/02/16 by Md. Mohi Uddin Khan, Khan, Md. Mohi Uddin, Abdullah Bin Shams +3 · 1 citation
Computer Science · Engineering · #Artificial Intelligence (cs.AI) #Context-Aware Activity Recognition Systems #FOS: Computer and information sciences #FOS: Electrical engineering #Indoor and Outdoor Localization Technologies #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Signal Processing (eess.SP) #Wireless Networks and Protocols #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2202.08146
openalex publication_date 2022/02/16 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Human Activity Recognition (HAR) research has gained significant momentum due\nto recent technological advancements, artificial intelligence algorithms, the\nneed for smart cities, and socioeconomic transformation. However, existing\ncomputer vision and sensor-based HAR solutions have limitations such as privacy\nissues, memory and power consumption, and discomfort in wearing sensors for\nwhich researchers are observing a paradigm shift in HAR research. In response,\nWiFi-based HAR is gaining popularity due to the availability of more\ncoarse-grained Channel State Information. However, existing WiFi-based HAR\napproaches are limited to classifying independent and non-concurrent human\nactivities performed within equal time duration. Recent research commonly\nutilizes a Single Input Multiple Output communication link with a WiFi signal\nof 5 GHz channel frequency, using two WiFi routers or two Intel 5300 NICs as\ntransmitter-receiver. Our study, on the other hand, utilizes a Multiple Input\nMultiple Output radio link between a WiFi router and an Intel 5300 NIC, with\nthe time-series Wi-Fi channel state information based on 2.4 GHz channel\nfrequency for mutual human-to-human concurrent interaction recognition. The\nproposed Self-Attention guided Bidirectional Gated Recurrent Neural Network\n(Attention-BiGRU) deep learning model can classify 13 mutual interactions with\na maximum benchmark accuracy of 94% for a single subject-pair. This has been\nexpanded for ten subject pairs, which secured a benchmark accuracy of 88% with\nimproved classification around the interaction-transition region. An executable\ngraphical user interface (GUI) software has also been developed in this study\nusing the PyQt5 python module to classify, save, and display the overall mutual\nconcurrent human interactions performed within a given time duration. ...\n