Current Result Document :
| 한글제목(Korean Title) |
Transportation Mode Detection for People with Mobility Disabilities: A Pilot Study on a User-independent Model |
| 영문제목(English Title) |
Transportation Mode Detection for People with Mobility Disabilities: A Pilot Study on a User-independent Model |
| 저자(Author) |
Sungjin Hwang
Jiwoong Heo
Jucheol Moon
Hansung Kim
Jaehyuk Cha
Kwanguk (Kenny) Kim
|
| 원문수록처(Citation) |
VOL 48 NO. 02 PP. 0983 ~ 0985 (2021. 12) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
Transportation mode detection (TMD) can improve our daily lives by understanding the mobility patterns of people. Moreover, such understanding of the mobility can be especially beneficial to wheelchair users. However, the previous study showed that TMD performance decreased when tested with new users which were not used for machine learning, and this problem also possible on TMD for people with mobility disabilities. Therefore, we investigate the performance of TMD for wheelchair users (wTMD) on new users, and suggest a novel method to improve the performance on new users. Our results show that wTMD performance toward new users decreased, and proposed method (DenseNetbased-model) is superior than previous method (CNN-based-model) on new users. It indicates the importance of an evaluation method and deep CNN with good information exchange for the generalization of wTMD on new users. |
| 키워드(Keyword) |
|
| 파일첨부 |
PDF 다운로드
|
