PULSE: Personalized Opportunistic Learning for Wearable Devices

Mattia Giovanni Campana; Dariush Salami; Hüseyin Yiğitler; Stephan Sigg

cs.LGOct 20, 2026

PULSE: Personalized Opportunistic Learning for Wearable Devices

Authors: Mattia Giovanni Campana, Dariush Salami, Hüseyin Yiğitler, Stephan Sigg

Organizations: IIT-CNR · Aalto University

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.

Figures & tables

Dataset	Scenario	Users	Samples	Labels (Max. %)
UCI-HAR	HAR	30	10.32 K	6 (18.9)
HHAR	HAR	9	87.6 K	6 (19.5)
SisFall	FD	38	55.1 K	2 (67.4)
WESAD	SR	15	1 K	3 (54.5)

TABLE I : Summary of the evaluation datasets.

Panel of Figure 2 — Fig. 1 : Pretraining curves for the task-specific models.

Fig. 2 : Temporal evolution of contact opportunities.

		Cambridge			Community–Dense			Community–Sparse			Edge-Markov–Dense			Edge-Markov–Sparse
Scenario	Method	Loss	Acc.	F1	Loss	Acc.	F1	Loss	Acc.	F1	Loss	Acc.	F1	Loss	Acc.	F1
HAR	Local	.40	.87	.86	.88	.68	.64	.89	.68	.64	.87	.68	.64	.88	.68	.64
	OppAVG	.41	.86	.86	.96	.61	.56	.97	.61	.56	.96	.61	.55	.97	.61	.55
	OppFL	1.05	.55	.52	.91	.65	.64	1.24	.54	.51	.94	.65	.64	1.20	.54	.51
	OPTIMIST	.91	.59	.57	.86	.70	.69	.95	.65	.63	.83	.71	.70	1.00	.65	.63
	PULSE	.39	.87	.87	.55	.82	.82	.71	.77	.76	.58	.82	.81	.69	.76	.75

TABLE II : Final test performance. Bold and underlined values indicate the best and second-best available methods, respectively, for each dataset, contact-trace configuration, and metric.

Scenario	OppAVG	OppFL	OPTIMIST	PULSE
HAR–C	10.99	10.99	5.31	1.42
HAR–CD	33.35	33.35	18.52	3.06
HAR–CS	6.66	6.66	5.90	1.87
HAR–EMD	31.20	31.20	15.86	3.13
HAR–EMS	5.79	5.79	5.04	2.42
FD–C	14.81	11.93	7.35	1.50

TABLE III : Average number (

\times 1000

) of completed peer-assisted updates across the simulation.

Fig. 3 : Exploded view of the SensWear ring configuration: enclosure, flexible main board, curved battery, and interchangeable daughter boards.

Panel of Figure 7 — Fig. 4 : On-device feasibility of PULSE. (a) Energy distributions for local, initiator, and responder operations; (b) static memory footprint of the PULSE implementation; and (c) latency distributions measured on a SensWear device pair. Energy roles were measured separately and therefore do not represent paired exchanges.

Personalized Learning

See all 18 papers

May 22, 2026

Federated Learning over Human-Body Communication for On-Body Edge Intelligence: A Survey, Taxonomy, and BODYFED-HBC Scheduling Vignette

Human-body communication (HBC) is a promising physical substrate for wearable body-area networks because it can localize communication around the body and reduce the burden of conventional radio links. Federated learning (FL) is a promising learning substrate because it can reduce raw-data centralization for physiological and behavioral sensing. Yet these two literatures remain weakly connected: FL for wearables usually abstracts the communication layer, whereas HBC research usually abstracts learning and model-update traffic. This article surveys the intersection of HBC, wireless body-area networks, wearable FL, Internet-of-Bodies privacy, and edge-intelligence optimization. We propose a taxonomy that distinguishes intra-body, body-hub, cross-user, and clinical-cloud FL deployments, and we identify the open problem of body-channel-aware FL: learning protocols whose client selection, update compression, and aggregation are controlled by posture-dependent HBC links, residual energy, sensor memory, and privacy risk. To make the research agenda concrete, we introduce BODYFED-HBC as a reference architecture and provide an optimization formulation and scheduling algorithm. We further specify a reproducible simulation vignette that combines public wearable datasets with empirical body-coupled-communication signal-loss models. The article concludes with open datasets, evaluation metrics, limitations, and research directions for computer scientists working above the hardware layer.

Koffka Khan

Department of Computing and Information Technology The University of the West Indies, St. Augustine, Trinidad and Tobago

Client Selection in Federated Learning Federated Learning

Ask PDF

Jun 28, 2026

SP-CACW: Convergence-Aware Client Weighting for Selfish Personalized Learning

Collaborative learning is sustainable only when it benefits each participant. Standard federated learning optimizes a global average objective, which can under perform for clients whose data distributions differ substantially from the population. We study selfish personalization: how a designated target client can use peer gradients to minimize its own risk while avoiding negative transfer. We propose SP-CACW, a convergence-aware client-weighting framework that selects aggregation weights by minimizing an upper bound on the target client's convergence error. The resulting rule explicitly trades off peer bias against stochastic variance and can assign zero weight to harmful peers. We provide convergence guarantees under smoothness and bounded-variance assumptions and evaluate the method on MNIST, CIFAR-100, and LEAF Shakespeare, where it is competitive with or improves over strong personalized and clustering baselines.

Yaron Kiselman, Kfir Y. Levy

Technion Haifa, Israel

Federated Learning Aggregation Non-IID Federated Learning Federated Learning

Ask PDF

Oct 10, 2026

Data-Topology Compatibility Enables Personalization in Decentralized Learning

Decentralized learning must balance collaboration with personalization across heterogeneous agents. This paper studies when cooperation can support statistically coherent groups of personalized models. We show that data-topology compatibility enables robust orbit-wise personalization: when local objectives and communication topology have (approximately) compatible automorphisms, agents within an orbit share (approximately) equal limiting distributions, while different orbits remain distinct without explicit personalization. This creates a new intermediate regime between global consensus and independent local learning. Under exact data-topology compatibility, as a reference system, we decompose the stochastic learning dynamics into orbit-level and transverse components, derive a necessary and sufficient stability criterion for distributional synchronization, and characterize the orbit-structured stationary law. Modelling an approximate data-topology compatible system as an

\varepsilon

-perturbed reference system, we prove the introduced errors have uniform-in-time and stationary

O(\varepsilon)

bounds in 2-Wasserstein distance. We also prove sufficient conditions for lower stationary average excess risk than both independent local training and the best single shared parameter. Experiments on synthetic data, MNIST, and CIFAR-10 are in full agreement with the theory. The code is available at https://github.com/Veronicaaa1/orbit-personalization.

Veronica Xu, Lingfei Wang, Chenlei Leng +1

Decentralized Learning Distributed Optimization

Ask PDF

Abstract

Personalized Learning

Explore similar work

Federated Learning over Human-Body Communication for On-Body Edge Intelligence: A Survey, Taxonomy, and BODYFED-HBC Scheduling Vignette

SP-CACW: Convergence-Aware Client Weighting for Selfish Personalized Learning

Data-Topology Compatibility Enables Personalization in Decentralized Learning