Beyond 1 × 1 Convolutions: A Dynamic Select-and-Fuse Channel Sampling Strategy for On-Device Human Activity Recognition
- 5 January 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (7) , 14923-14939
Abstract
The proliferation of low-cost, portable sensors has made wearable human activity recognition (HAR) a cornerstone for real-time health monitoring and behavior analysis. However, deploying accurate yet lightweight deep learning models on resource-constrained wearable devices poses a significant challenge for on-device activity recognition. While channel pruning is a common solution to accelerate deep convolutional neural networks (CNNs), existing works often require specialized implementations or pretrained models, which potentially degrade performance by simply removing an entire channel, limiting their ability to handle complex multimodal sensor inputs. Moreover, lightweight CNN design, particularly the heavy use of $1\times 1$ convolution layers for channel squeezing, remains inefficient for sensor-based HAR, which consumes resources without expanding the receptive field due to their pointwise nature. To address these issues, we propose a novel dynamic channel sampling module, select-and-fuse (SaF), specifically designed for sensor-based HAR. SaF divides channels into subsets and performs a dynamic, input-dependent selection from them, with the picking decision being made per time step based on the input sensor signal activations, allowing for fine-grained feature adaptation to multimodal sensor signals. While integrated into compact backbones, SaF significantly reduces model size and inference latency while maintaining high accuracy. Extensive evaluations on public UCI-HAR, OPPORTUNITY, WISDM, and UniMiB-SHAR benchmarks confirm a favorable performance–cost tradeoff. Crucially, we measure actual inference latency on a Raspberry Pi, proving its practicality for resource-constrained HAR applications. The code will be released.Keywords
Funding Information
- National Natural Science Foundation of China (62373194)
- Cultivating Plan Program for the Leader in Science and Technology of Yunnan Province, China (202005AC160005)
- Ten Thousand Talent Plans for Young of Yunnan Province, China (YNWR-QNBJ-2019-188)
This publication has 42 references indexed in Scilit:
- CNNs, RNNs and Transformers in human action recognition: a survey and a hybrid modelArtificial Intelligence Review, 2025
- Fixing deep early exit ensembles for sensor-based human activity recognition through uncertainty quantificationApplied Soft Computing, 2025
- MobileNetV4: Universal Models for the Mobile Ecosystem2024
- A review of convolutional neural networks in computer visionArtificial Intelligence Review, 2024
- Sensor-based and vision-based human activity recognition: A comprehensive surveyPattern Recognition, 2020
- Smartphone and Smartwatch-Based Biometrics Using Activities of Daily LivingIEEE Access, 2019
- ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design2018
- AMC: AutoML for Model Compression and Acceleration on Mobile Devices2018
- Human activity recognition with smartphone sensors using deep learning neural networksExpert Systems with Applications, 2016
- Human Activity Recognition on Smartphones Using a Multiclass Hardware-Friendly Support Vector Machine2012