Original title: Intelligent Knee Sleeves: A Real-time Multimodal Dataset for 3D Lower Body Motion Estimation Using Smart Textile
Authors: Wenwen Zhang, Arvin Tashakori, Zenan Jiang, Amir Servati, Harishkumar Narayana, Saeid Soltanian, Rou Yi Yeap, Meng Han Ma, Lauren Toy, Peyman Servati
In their article, researchers introduce the Texavie MarsWear Knee Sleeves, offering insights into human movement dynamics. They present a diverse dataset obtained from these intelligent sleeves, synchronized with visual motion capture data, to construct 3D human models. This camera-free system, using machine learning, accurately predicts lower body joints during various activities, even those not previously recorded. The system exhibits an average 7.21-degree error across eight joints, showcasing its reliability beyond just knees. This innovation enables seamless human pose estimation, transcending visual limitations. The findings highlight the potential of wearable sensing in diverse fields, from fitness and sports to healthcare and rehabilitation, emphasizing movement and pose estimation’s broad applicability.
Original article: https://arxiv.org/abs/2311.12829