松井 智一
生駒 : 奈良先端科学技術大学院大学, 2023.6
授業アーカイブRecognition of in-home activities is essential for predicting electricity demand and monitoring services for the elderly. In order to realize in-home activity recognition not only in smart homes but also in ordinary households, a sensing system that is easy to install and a suitable activity recognition method are required. This presentation describes our sensing system, which is easy to install, and an activity recognition method for multi-person environments. Experiments show that the proposed sensing system can operate maintenance-free for an extended period, is helpful for data collection, and that the proposed activity recognition method improves the accuracy of multi-person activity recognition.
2023
情報科学領域・コロキアム ; 2023年度
講演者所属:情報科学領域
講演日:2023年6月20日 3限
講演場所:情報科学棟 エーアイ大講義室(L1)
英語 (eng)
英語 (eng)