Multi-task Self-Supervised Learning for Human Activity Detection

计算机科学 人工智能 杠杆(统计) 特征学习 学习迁移 机器学习 卷积神经网络 监督学习 深度学习 任务(项目管理) 标记数据 二元分类 无监督学习 半监督学习 模式识别(心理学) 人工神经网络 支持向量机 经济 管理
作者
Aaqib Saeed,Tanir Ozcelebi,Johan J. Lukkien
出处
期刊:Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies [Association for Computing Machinery]
卷期号:3 (2): 1-30 被引量:128
标识
DOI:10.1145/3328932
摘要

Deep learning methods are successfully used in applications pertaining to ubiquitous computing, health, and well-being. Specifically, the area of human activity recognition (HAR) is primarily transformed by the convolutional and recurrent neural networks, thanks to their ability to learn semantic representations from raw input. However, to extract generalizable features, massive amounts of well-curated data are required, which is a notoriously challenging task; hindered by privacy issues, and annotation costs. Therefore, unsupervised representation learning is of prime importance to leverage the vast amount of unlabeled data produced by smart devices. In this work, we propose a novel self-supervised technique for feature learning from sensory data that does not require access to any form of semantic labels. We learn a multi-task temporal convolutional network to recognize transformations applied on an input signal. By exploiting these transformations, we demonstrate that simple auxiliary tasks of the binary classification result in a strong supervisory signal for extracting useful features for the downstream task. We extensively evaluate the proposed approach on several publicly available datasets for smartphone-based HAR in unsupervised, semi-supervised, and transfer learning settings. Our method achieves performance levels superior to or comparable with fully-supervised networks, and it performs significantly better than autoencoders. Notably, for the semi-supervised case, the self-supervised features substantially boost the detection rate by attaining a kappa score between 0.7-0.8 with only 10 labeled examples per class. We get similar impressive performance even if the features are transferred from a different data source. While this paper focuses on HAR as the application domain, the proposed technique is general and could be applied to a wide variety of problems in other areas.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
PDF的下载单位、IP信息已删除 (2025-6-4)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
1秒前
弹剑作歌完成签到,获得积分10
6秒前
jy发布了新的文献求助10
6秒前
Lucky发布了新的文献求助10
7秒前
落榜美术生完成签到,获得积分10
8秒前
8秒前
9秒前
飘逸的鸽子完成签到,获得积分10
12秒前
XL神放发布了新的文献求助10
12秒前
祗想静静嘚完成签到 ,获得积分10
13秒前
Gjj发布了新的文献求助10
13秒前
Lucky完成签到,获得积分10
17秒前
迷你的无声完成签到,获得积分10
19秒前
20秒前
刀锋发布了新的文献求助10
25秒前
Akim应助Arya123000采纳,获得10
30秒前
iorpi完成签到,获得积分10
32秒前
谷粱安卉完成签到 ,获得积分10
32秒前
称心的战斗机关注了科研通微信公众号
35秒前
快乐的易巧完成签到,获得积分10
37秒前
人间冒险发布了新的文献求助10
37秒前
Gjj完成签到,获得积分10
39秒前
44秒前
郭WL完成签到,获得积分10
45秒前
CipherSage应助guard采纳,获得10
45秒前
阔达碧空发布了新的文献求助10
47秒前
Rondab应助meng采纳,获得10
47秒前
yh关闭了yh文献求助
47秒前
慕青应助刀锋采纳,获得10
48秒前
48秒前
李健的小迷弟应助容cc采纳,获得10
52秒前
英勇的绿海完成签到,获得积分10
54秒前
Rondab应助瘦瘦从梦采纳,获得10
59秒前
缘分完成签到 ,获得积分10
1分钟前
李爱国应助科研通管家采纳,获得10
1分钟前
科目三应助科研通管家采纳,获得10
1分钟前
1分钟前
领导范儿应助科研通管家采纳,获得10
1分钟前
1分钟前
1分钟前
高分求助中
The Mother of All Tableaux: Order, Equivalence, and Geometry in the Large-scale Structure of Optimality Theory 3000
Social Research Methods (4th Edition) by Maggie Walter (2019) 1030
A new approach to the extrapolation of accelerated life test data 1000
Indomethacinのヒトにおける経皮吸収 400
基于可调谐半导体激光吸收光谱技术泄漏气体检测系统的研究 370
Phylogenetic study of the order Polydesmida (Myriapoda: Diplopoda) 370
Robot-supported joining of reinforcement textiles with one-sided sewing heads 320
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 冶金 细胞生物学 免疫学
热门帖子
关注 科研通微信公众号,转发送积分 3994236
求助须知:如何正确求助?哪些是违规求助? 3534710
关于积分的说明 11266276
捐赠科研通 3274624
什么是DOI,文献DOI怎么找? 1806413
邀请新用户注册赠送积分活动 883273
科研通“疑难数据库(出版商)”最低求助积分说明 809731