HARMamba: Efficient Wearable Sensor Human Activity Recognition Based on Bidirectional Selective SSM

arXiv:2403.20183v1 Announce Type: new
Abstract: Wearable sensor human activity recognition (HAR) is a crucial area of research in activity sensing. While transformer-based temporal deep learning models have been extensively studied and implemented, their large number of parameters present significant challenges in terms of system computing load and memory usage, rendering them unsuitable for real-time mobile activity recognition applications. Recently, an efficient hardware-aware state space model (SSM) called Mamba has emerged as a promising alternative. Mamba demonstrates strong potential in long …

abstract arxiv challenges computing cs.ai cs.cv deep learning human memory parameters recognition research sensing sensor ssm temporal terms transformer type wearable