稳健性(进化)
计算机科学
跟踪(教育)
跟踪系统
计算机视觉
人工智能
实时计算
卡尔曼滤波器
心理学
教育学
生物化学
化学
基因
作者
Xiangyu Xu,Hao Kong,Jiadi Yu
标识
DOI:10.1145/3603165.3607428
摘要
With the rapid growth of artificial intelligence-of-things (AIoT) in recent years, the market for 3D human posture tracking has expanded to encompass a wide range of application scenarios, including various VR/AR/MR applications. However, current mainstream solutions for posture tracking, which rely on vision-based methods, suffer from concerns regarding privacy leakage and are heavily dependent on favorable lighting conditions. Conversely, although RF-based posture-tracking solutions have been proposed in recent years to address privacy issues and handle varying lighting conditions, doubts have been raised regarding their robustness. To push the limit of 3D human posture tracking, we present a mmWave-based robust multi-user 3D posture tracking system, which leverages a single commercial off-the-shelf (COTS) mmWave radar to track multiple users’ 3D postures simultaneously and keep the tracking result robust in real-world scenarios.
科研通智能强力驱动
Strongly Powered by AbleSci AI