文章基本信息

标题：A Benchmark Dataset for Depth Sensor Based Activity Recognition in a Manufacturing Process
本地全文：下载
作者：Don J. Rude ; Don J. Rude ; Stephen Adams 等
期刊名称：IFAC PapersOnLine
印刷版ISSN：2405-8963
出版年度：2015
卷号：48
期号：3
页码：668-674
DOI：10.1016/j.ifacol.2015.06.159
语种：English
出版社：Elsevier
摘要：Abstract Algorithms for automated recognition of human activities are crucial for supporting the next generation of process measures in manufacturing. While there is active research underway for many sensor systems and algorithms they will need to be tested in real-world conditions in order to mature and become robust or generalized enough for broad deployment in industry. In this paper we present a case study and dataset from a real-world setting along with three performance measures for six common classifiers. The intent is to provide a dataset and baseline performance level metrics so that others may compare their activity recognition algorithms to a common standard.
关键词：KeywordsActivity recognitionmachine learningdepth cameraKinectdata setsperformance analysis