DMSCNet: A Dilated Multi-Scale Contrastive Attention Network for Sensor-Based Human Activity Recognition
Applied Sciences · 15.06.2026 · 16 (12) · 6037
DOI: 10.3390/app16126037 ↗Kurz gefasst
Fachartikel in der Zeitschrift Applied Sciences, online erschienen am 15.06.2026. Verfasst von Qingshan Wu und 3 weiteren Autor:innen. Die Arbeit nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Natural Science Foundation of Shandong Province of China.
Abstract (Originalsprache)
Wearable-sensor human activity recognition (HAR) plays a key role in health monitoring, elderly care, and human–computer interaction. Deep learning dominates the field, but two limitations remain. CNNs with fixed kernels cannot capture cross-scale temporal events such as gait cycles and postural transitions in a single layer, and softmax attention on small sensor datasets is often diluted by common-mode background responses across the sequence. We propose DMSCNet, an end-to-end framework with two modules. The Dilated Multi-Scale Branch Block (DMSB) combines a shared bottleneck, parallel dilated convolutions, a pooling bypass, and SE-based channel recalibration to widen the temporal receptive field under a controlled parameter budget. The Contrastive Temporal Attention (CTA) module adopts a dual-path differential design, in which the two paths learn overlapping but non-identical attention patterns and their subtraction suppresses shared low-level responses while preserving the discriminative positions each path locks onto, encoded with opposite signs. DMSB and CTA are cascaded into a DMSC Block and stacked residually. On UCI-HAR, USC-HAD, and RealWorld, DMSCNet reaches F1-scores of 97.65%, 91.80%, and 99.05%, outperforming nine baselines. Ablations confirm that SE acts along the channel axis and CTA along the temporal axis, and visualization reveals a dynamic–static dichotomy together with a signed bipolar encoding pattern produced by the dual-path subtraction.
Förderung
- National Natural Science Foundation of China
- Natural Science Foundation of Shandong Province of China
Zitieren
APA
Wu, Q., Chu, S., Li, K., & Wang, L. (2026). DMSCNet: A Dilated Multi-Scale Contrastive Attention Network for Sensor-Based Human Activity Recognition. Applied Sciences, 16(12), 6037. https://doi.org/10.3390/app16126037
Vancouver
Wu Q, Chu S, Li K, Wang L. DMSCNet: A Dilated Multi-Scale Contrastive Attention Network for Sensor-Based Human Activity Recognition. Applied Sciences. 2026;16(12):6037. doi:10.3390/app16126037
BibTeX
@article{Wu2026,
author = {Wu, Qingshan and Chu, Shengguang and Li, Kewen and Wang, Liechong},
title = {DMSCNet: A Dilated Multi-Scale Contrastive Attention Network for Sensor-Based Human Activity Recognition},
journal = {Applied Sciences},
year = {2026},
volume = {16},
number = {12},
pages = {6037},
doi = {10.3390/app16126037}
}
EUVIMED