2019 Fiscal Year Final Research Report
Sophisticated indoor context estimation based on real world data
| Project/Area Number |
17H01762
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Perceptual information processing
|
|---|
| Research Institution | Nagoya University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2017-04-01 – 2020-03-31
|
| Keywords | コンテキスト推定 / 行動認識 / 深層学習 / 歩行者自律測位 / 屋内測位 / 回転磁石マーカ / PDR / 行動推定 |
|---|
| Outline of Final Research Achievements |
In this study, we examined a framework for integrating activity recognition and indoor positioning, and in particular, we aimed to make use of deep learning.
In PDR (Pedestrian Dead-Reckoning), we constructed an end-to-end deep learning PDR that directory acquires a relative position change from sensor data. We also propose indoor context estimation using environmentally installed devices. We use noise reduction based on deep learning for BLE (Bluetooth Low Energy) radio strength. Our data obtained through the research activities were made available to the world via http://hub.hasc.jp. In addition, we have developed a system named Synerex for supply and demand exchange as a fundamental platform for data collection.
|
| Free Research Field |
知覚情報処理
|
| Academic Significance and Societal Importance of the Research Achievements |
屋内コンテキスト推定技術は，ユーザの行動意図を理解し，支援することが究極の目的と言える．本研究はその基盤技術の獲得を目指しており，「屋内コンテキスト推定」という新しい普遍的な研究分野の確立が期待でき，その社会適用範囲も非常に広い．また代表者はこれまですでに国際ワークショップ(HASCA2013-2020)，国際コンペティション(PDR Challenge)を実現しており，本研究により日本の「コンテキスト・アウェア」研究技術を世界に広めることができた。さらに、End-to-endの深層学習による成果や、開発された基盤システム Synerex は多様な活用可能性が期待できる。
|
