On the Metrics and Adaptation Methods for Domain Divergences of sEMG-based Gesture Recognition

TL;DR

This paper introduces a novel metric for quantifying domain divergence and a two-stage domain adaptation method using an RNN-based architecture, specifically addressing challenges in sEMG gesture recognition across different sessions and subjects.

Contribution

It presents a new probability distribution-based, transductive metric and a novel domain adaptation approach tailored for time-series classification in sEMG gesture recognition.

Findings

The proposed metric effectively measures domain divergence without source data samples.

The adaptation method improves gesture recognition accuracy across domain shifts.

Experimental results demonstrate enhanced performance in inter-session and inter-subject scenarios.

Abstract

We propose a new metric to measure domain divergence and a new domain adaptation method for time-series classification. The metric belongs to the class of probability distributions-based metrics, is transductive, and does not assume the presence of source data samples. The 2-stage method utilizes an improved autoregressive, RNN-based architecture with deep/non-linear transformation. We assess our metric and the performance of our model in the context of sEMG/EMG-based gesture recognition under inter-session and inter-subject domain shifts.