Projects
A list of projects I have participated in
Improving lab-to-field generalization of activity recognition models
(Postdoc) Activity recognition models show high performance when trained and tested with data collected in controlled environments like experiments. However, when used in real-life, the performance of these models tends to drop significantly. We study two causes for this: new user scenario and usage of different devices. We propose solutions based on feature learning.
Featured Publications:
- Paula Lago, Moe Matsuki, Kohei Adachi, and Sozo Inoue. 2021. Using additional training sensors to improve single-sensor complex activity recognition. 2021 International Symposium on Wearable Computers. Association for Computing Machinery, New York, NY, USA, 18–22.
- Lago P., Takeda S., Okita T., Inoue S. (2019) MEASURed: Evaluating Sensor-Based Activity Recognition Scenarios by Simulating Accelerometer Measures from Motion Capture. In: Kawaguchi N., Nishio N., Roggen D., Inoue S., Pirttikangas S., Van Laerhoven K. (eds) Human Activity Sensing. Springer Series in Adaptive Environments. Springer, Cham
- P. Lago and S. Inoue, Comparing Feature Learning Methods for Human Activity Recognition: Performance study in new user scenario, 2019 Joint 8th International Conference on Informatics, Electronics & Vision (ICIEV) and 2019 3rd International Conference on Imaging, Vision & Pattern Recognition (icIVPR), Spokane, WA, USA, 2019, pp. 118-123.
Context-enriched patterns of behavior at home for Ambient Assisted Living
(PhD project) In Ambient Assisted Living routine deviations are usually used as a sign that assistance is needed. However, to reduce false alarms, context must be considered to assess when a routine deviation is truly an anomaly or when it is a result of a change in context, i.e. it is raining and the person can’t go out. In this project, I proposed context-enriched behavior patterns to describe routines with the context they occur on and a stream learning algorithm to discover them.
Featured Publications and Datasets
- Paula Lago, Claudia Roncancio, Claudia Jiménez-Guarín. Learning and managing context enriched behavior patterns in smart homes. Future Generation Computer Systems, Volume 91,2019, Pages 191-205.
- The ContextAct@A4H Real-Life Dataset of Daily-Living Activities
Automatic Nurse Care Record Recreation with Activity Recognition
(Postdoc at Kyutech) By using activity recognition based on sensor data and historical data of nurses activities, we aim at reducing documentation workload by partially filling the nurse’s record for the day.
Featured Publications and Datasets
- Inoue, S., Lago, P., Hossain, T., Mairittha, T., & Mairittha, N. (2019). Integrating Activity Recognition and Nursing Care Records: The System, Deployment, and a Verification Study. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 3(3), 1-24.
- Nursing activities dataset