PREFIXKD-EDGE: REAL-TIME LOCOMOTION-INTENT PREDICTION WITH PREFIX SELF-DISTILLATION ON LIGHTWEIGHT TEMPORAL NETWORKS
Innovation and Technology Studies · 12.01.2024 · 1 (1) · 30-38
DOI: 10.61784/its4025 ↗Kurz gefasst
Fachartikel in der Zeitschrift Innovation and Technology Studies, online erschienen am 12.01.2024. Verfasst von WenBin Shang und 1 weiteren Autor:innen. Die Arbeit wurde bisher 12-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Edge intent inference must make useful decisions before a sensing window is complete, while respecting tight memory and latency budgets. This paper studies early locomotion-mode prediction from wearable inertial sequences and introduces PrefixKD-Edge, a deployment-oriented framework that combines a compact temporal convolutional backbone, training-only full-window-to-prefix self-distillation, and confidence-gated decision timing. The task is evaluated on the public UCI Human Activity Recognition Using Smartphones dataset using its official subject-disjoint split. We retain three dynamic classes—level walking, stair ascent, and stair descent—and use the six supplied body-acceleration and gyroscope sequence channels rather than the 561 engineered features. A causal four-sample block mean reduces 50-Hz windows to 12.5 Hz, after which predictions are tested at 25%, 50%, 75%, and 100% observation. Across three random seeds, the 7,907-parameter PrefixKD-CNN attains macro-F1 scores of 94.30±1.76%, 97.64±1.62%, 98.00±1.43%, and 98.13±1.33%, respectively. Its measured FP32 TorchScript batch-1 median latency is 0.031–0.034 ms on one server CPU thread. With a confidence threshold of 0.90, the adaptive policy reaches 98.20±1.18% accuracy while observing 42.72±1.97% of the window on average (1.09 s). The same-backbone distillation gain is positive but small, so no statistical-significance claim is made. The results show that hardware-friendly backbones, prefix-aware training, and explicit decision policies can provide a practical accuracy–earliness operating point without adding inference parameters.
Zitieren
APA
Shang, W., & Chen, T. (2024). PREFIXKD-EDGE: REAL-TIME LOCOMOTION-INTENT PREDICTION WITH PREFIX SELF-DISTILLATION ON LIGHTWEIGHT TEMPORAL NETWORKS. Innovation and Technology Studies, 1(1), 30-38. https://doi.org/10.61784/its4025
Vancouver
Shang W, Chen T. PREFIXKD-EDGE: REAL-TIME LOCOMOTION-INTENT PREDICTION WITH PREFIX SELF-DISTILLATION ON LIGHTWEIGHT TEMPORAL NETWORKS. Innovation and Technology Studies. 2024;1(1):30-38. doi:10.61784/its4025
BibTeX
@article{Shang2024,
author = {Shang, WenBin and Chen, TingJie},
title = {PREFIXKD-EDGE: REAL-TIME LOCOMOTION-INTENT PREDICTION WITH PREFIX SELF-DISTILLATION ON LIGHTWEIGHT TEMPORAL NETWORKS},
journal = {Innovation and Technology Studies},
year = {2024},
volume = {1},
number = {1},
pages = {30--38},
doi = {10.61784/its4025}
}
EUVIMED