基于圆形空洞卷积神经网络 CDCNN 的智能鞋垫活动识别方法
Activity Recognition from Smart Insole Sensor Data Using a Circular Dilated CNN
AI 导读
研究者提出一种基于圆形空洞卷积神经网络（CDCNN）的智能鞋垫活动分类系统，处理压力、加速度计和陀螺仪多模态时序数据。模型使用 160 帧窗口、24 通道输入，在 Stand/Walk/Sit/Tandem 四分类任务上，跨被试评估取得 86.42% 测试准确率。该方法可用于智能鞋垫设计、健康监测与康复领域。
正文 · 原文
Abstract:Smart insoles equipped with pressure sensors, accelerometers, and gyroscopes offer a non-intrusive way of monitoring human gait and posture. We present an activity classification system based on a circular dilated convolutional neural network (CDCNN) that processes multi-modal time-series data from such insoles. The model operates on 160-frame windows with 24 channels (18 pressure, 3 accelerometer, 3 gyroscope axes), achieving 86.42% test accuracy in a subject-independent evaluation on a four-class task (Standing, Walking, Sitting, Tandem). Permutation feature importance reveals the distribution of contributions made by various sensors to the classification process. The proposed approach has potential applications in areas such as smart insole design, health monitoring, and rehabilitation.
| Comments: | Published in Current Directions in Biomedical Engineering |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2603.04477 [cs.LG] |
| (or arXiv:2603.04477v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2603.04477 arXiv-issued DOI via DataCite |
|
| Journal reference: | Current Directions in Biomedical Engineering (2026) |
| Related DOI: | https://doi.org/10.1515/cdbme-2026-0267
DOI(s) linking to related resources |
Submission history
From: Yanhua Zhao [view email]
[v1]
Wed, 4 Mar 2026 15:27:14 UTC (1,281 KB)
[v2]
Sat, 24 Oct 2026 19:55:43 UTC (1,076 KB)
来源：arXiv cs.AI · arxiv.org