A Dining Context-Aware System with Mobile and Wearable Devices

Presentation on theme: "A Dining Context-Aware System with Mobile and Wearable Devices"— Presentation transcript:

1 A Dining Context-Aware System with Mobile and Wearable Devices
2015 International Symposium on Advances in Pervasive Computing and Intelligence Systems Let the presentation started. In this paper, we propose the recognition system that recognizes whether the user is eating or not, using mobile and wearable device. (20초 이내) K. H. Kim and S. B. Cho Softcomputing Lab, Univ. of Yonsei

2 발표 순서 Description of the problem Related works The proposed method
Experiments Conclusion and future works Here is the outline of the presentation. First, I give a brief description of the problem to solve. This part includes what dinning and wearable device means, why recognizing this activity is useful, etc. Second part shows the related works about human activity recognition and dynamic Bayesian network. And next, I’ll gonna explain how we have tried to solve this problem using Bayesian network, and experiments for showing the usefulness of the proposed method. Conclusion and future works are followed lastly. (1분 이내)

3 발표 장면 및 공간

4 질의 응답 왜 식사행위를 인식합니까? 헬스케어 서비스, 고수준의 상황을 인지하기 위한 접근법 제시 식사행위가 다양한 맥락에서 발생할 수 있어 인식이 어렵다고 하셨는데요, 실험 부분에 식사행위 각각에 대한 내부분석이 수행된 바가 있나요? 샌드위치 and 스테이크 but 데이터 양 부족 -> 진행중

5 학회 발표 Exploring Accuracy-Cost Tradeoff in In-Home Living Activity Recognition based on Power Consumptions and User Positions 문제 : Elderly monitoring 시스템 -> 유지비용, 개인정보 유출 등 문제 -> power meter & ultrasonic positioning sensor만으로 인식 방법 : Power meter의 평균값과 position sensor의 중간값 (각 16개) -> SVM, DT, RF 실험 : RF가 가장 우수, windo별/sensor별/행위별 얻은 것 The Statistic Bureau, Ministry of Internal Affairs and Communications, > 주요 일상행위를 3그룹으로 분류 (식사 -> 1st activity) 공간 묘사 그림 (다음 슬라이드)

6 학회 발표 : 공간 묘사 방법