Indoor Location Semantic Labeling with Mobile and Wearable Sensing

2021 Fiscal Year Final Research Report

• Project/Area Number: 17H04679
• Research Category: Grant-in-Aid for Young Scientists (A)
• Allocation Type: Single-year Grants
• Research Field: Information network
• Research Institution: Osaka University
• Principal Investigator: Maekawa Takuya 大阪大学, 情報科学研究科, 准教授 (50447025)
• Project Period (FY): 2017-04-01 – 2021-03-31
• Keywords: ユビキタスコンピューティング / ウェアラブルコンピューティング / 屋内位置推定 / 行動認識

Outline of Final Research Achievements

This study developed techniques for predicting semantic labels to various indoor locations. Specifically, we developed methods for estimating static and dynamic labels for indoor locations. When a user is in a bedroom at time t, for example, we estimate a class label for "bedroom" using sensor data at time t observed by a sensor device possessed by the user. We also proposed a method for estimating dynamic labels of indoor objects whose states change dynamically, such as doors and windows, using information such as the Doppler effect obtained from active sound sensing. These results were accepted as several full papers in UbiComp (ACM IMWUT), the top international conference in the field, and IEEE Sensors Journal.

Free Research Field

ユビキタスコンピューティング

Academic Significance and Societal Importance of the Research Achievements

「寝室」、「浴室」、「会議室」といったユーザが現在居る場所の意味が推定できれば、そのユーザの場所の意味に応じたコンテキストアウェアサービスが実現できる。例えば、見守り対象の高齢者がどのような意味を持つ場所に居るのかを、遠隔家族などに提示できる。また、ドアなどの屋内オブジェクトの状態が推定できれば、サービスロボットの移動計画生成や空調の自動調節などの応用に利用できる。