Selective peer learning cuts communication for wearable personalization
PULSE: Personalized Opportunistic Learning for Wearable Devices
Machine LearningArtificial IntelligenceDistributed, Parallel, and Cluster Computing
Summary
Wearable devices that track health need to adapt to new data after they are sold, but devices often have limited data on their own and poor or inconsistent connectivity. The authors created PULSE, a way for wearable devices to learn from nearby devices only when it helps them improve their own health tracking. This method chooses when and who to learn from based on benefit and cost, saving energy and data transfer. Testing shows PULSE keeps or improves accuracy while greatly reducing communication and works well on actual wearable hardware.
What this means in practice
- •For wearable device designers: Create wearable health trackers that personalize using nearby devices only when it benefits accuracy and saves communication and energy costs.$Commercial implications: Enables marketable health wearables that adapt better in real-world conditions while minimizing battery and data costs through selective peer learning.
- •For mobile app developers: Build health monitoring apps that improve on-device learning by leveraging selective peer updates to maintain user privacy and reduce network use.
Authors
Mattia Giovanni Campana, Dariush Salami, Hüseyin Yiğitler, Stephan Sigg
Abstract
Wearable devices increasingly support personalized health and well-being applications, but post-deployment data can differ substantially from those used for training, while purely local adaptation may be limited by the data available on a single device during real-world operation. Conventional federated and decentralized learning rely on recurrent communication and stable connectivity, whereas existing opportunistic approaches often exploit available encounters without adequately considering their utility, compatibility, or communication cost. We propose PULSE, an event-driven framework for personalized opportunistic learning that treats each encounter as an online collaboration decision, selecting either no interaction or a promising nearby peer. Peer gradients are incorporated only when compatible with the local learning direction, while past interactions guide future peer selection. Across human activity recognition, fall detection, and stress recognition under synthetic and real contact traces, PULSE preserves or improves personalized classification performance, while reducing peer-assisted exchanges by 80-94% compared with opportunistic model averaging. On-device experiments further demonstrate its feasibility on wearable hardware: peer-assisted operations require limited energy, while the complete implementation occupies 149.8 KiB of RAM and 561.8 KiB of flash. The obtained results show that selective opportunistic collaboration can provide effective personalization while limiting communication and resource demands on constrained wearable devices.