🤖 AI Summary
This work addresses the challenge of continuously decoding hand kinematics from surface electromyographic (EMG) signals by proposing KinEMbed, a novel framework that introduces cross-modal contrastive learning to the regression task of mapping EMG to continuous hand motion. KinEMbed jointly trains an EMG feature encoder and a kinematic target encoder to learn embedding representations that preserve the geometric structure of joint angle space, without requiring kinematic signals during inference. Evaluated on the NinaPro DB8 dataset, KinEMbed significantly outperforms baseline methods—including PCA, PLS, autoencoders, and CEBRA—with particularly notable improvements in decoding performance for the most challenging degrees of freedom of the thumb.
📝 Abstract
Decoding hand kinematics from surface electromyography (EMG) is a core challenge in wearable biosignal processing with clinical relevance for prosthetic control and motor rehabilitation. Most representation learning approaches for EMG focus on discrete gesture classification, and few focus on continuous regression. We present KinEMbed, a cross-modal contrastive learning framework for hand kinematics regression that jointly trains dual encoders -- one for windowed EMG features and one for kinematic (joint angle) targets. The resulting embeddings inherit the geometric structure of the kinematic space without requiring kinematic signals at inference time. Evaluating on the NinaPro DB8 dataset that includes both able-bodied users and subjects with limb difference (N=11), KinEMbed outperforms PCA, PLS, autoencoder and contrastive (CEBRA) baselines on held-out sessions, with largest gains on the most challenging thumb degrees of articulation. We position this work as a first step toward contrastive representation learning for regression of hand kinematics from structured wearable biosignals.