A Gait Data Compression and Reconstruction Framework for Edge Device using Low-Dimensional Attention Model with Autoencoder
2025 IEEE International Symposium on Circuits and Systems (ISCAS 2025), 2025
Overview
An attention-based autoencoder for learning compact representations of gait time-series data. The work examines the trade-off between compression and reconstruction fidelity, using a reduced-dimensional attention mechanism.