RelCon: Relative Contrastive Learning for a Motion Foundation Model for
Wearable Data
RelCon: Relative Contrastive Learning for a Motion Foundation Model for
Wearable Data
We present RelCon, a novel self-supervised *Rel*ative *Con*trastive learning approach that uses a learnable distance measure in combination with a softened contrastive loss for training an motion foundation model from wearable sensors. The learnable distance measure captures motif similarity and domain-specific semantic information such as rotation invariance. The learned distance …