A large-scale benchmark of 17 WHAR models across 30 datasets finds predictive performance has plateaued while efficiency favors compact neural models and random forests on the Pareto frontier.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.LG 2representative citing papers
ActiNet, consisting of a self-supervised 18-layer ResNet-V2 model followed by HMM smoothing, achieves mean macro F1 of 0.82 and Cohen's kappa of 0.86 on wrist accelerometer data from 151 participants, outperforming a random forest plus HMM baseline.
citing papers explorer
-
WHAR Arena: Benchmarking the State of the Art in Efficient Wearable Human Activity Recognition
A large-scale benchmark of 17 WHAR models across 30 datasets finds predictive performance has plateaued while efficiency favors compact neural models and random forests on the Pareto frontier.
-
ActiNet: An Open-Source Tool for Activity Intensity Classification of Wrist-Worn Accelerometry Using Self-Supervised Deep Learning
ActiNet, consisting of a self-supervised 18-layer ResNet-V2 model followed by HMM smoothing, achieves mean macro F1 of 0.82 and Cohen's kappa of 0.86 on wrist accelerometer data from 151 participants, outperforming a random forest plus HMM baseline.