![Professor Kim Kyung-joong, doctoral student Joo Ho-taek and Lee Seong-ha [Provided by Gwangju Institute of Science and Technology. Redistribution and retransmission prohibited.]](https://imgnews.pstatic.net/image/001/2024/07/31/AKR20240731049900054_04_i_P4_20240731095417665.jpg?type=w647)
Professor Kim Kyung-joong, doctoral student Joo Ho-taek and Lee Seong-ha
[Provided by Gwangju Institute of Science and Technology. Redistribution and retransmission prohibited.]
According to the news by Son Sang-won from Gwangju on the 31st, virtual reality (VR) technology that can enhance immersion without wearing special devices has been developed.
Gwangju Institute of Science and Technology (GIST) announced that a research team led by Professor Kim Kyung-joong of the Department of Convergence Technology developed a technology that analyzes the user’s foot pressure data using a high-resolution carpet-type touch sensor to recognize behavior.
To recognize and reflect the movements of VR users, sensors are essential, but wearing them on the body can cause discomfort or reduce immersion.
The research team developed a technology to recognize the user’s behavior by analyzing foot pressure data using carpet-type touch sensors laid on the floor without wearable devices.
To process foot pressure data, a new artificial intelligence model based on Vision Transformer (ViT) called Self-Training Vision Transformer (STViT) was applied to accurately distinguish similar movements.
ViT is a technology that analyzes small pieces of images to understand the entire image, mimicking the human visual system.
STVit of the research team is an autonomous learning method that continuously updates knowledge using the knowledge of previous learning stages, added to the existing ViT.
![Application of technology [Provided by Gwangju Institute of Science and Technology. Redistribution and retransmission prohibited.]](https://imgnews.pstatic.net/image/001/2024/07/31/AKR20240731049900054_05_i_P4_20240731095417670.jpg?type=w647)
Application of technology
[Provided by Gwangju Institute of Science and Technology. Redistribution and retransmission prohibited.]
Professor Kim Kyung-joong said, “It will be possible to use the interface to recognize the user’s movements more naturally and comfortably without wearable devices,” and expressed his expectation that it will solve the natural mobility issues in virtual reality and open up possibilities for providing a more immersive user experience.