The article presents a survey on transfer learning methods applied in the field of Human Activity Recognition (HAR). The authors highlight the challenges faced in sensor-based HAR due to the lack of large amounts of annotated data and varying real-world settings. The focus is on smart home and wearable applications of HAR. The survey provides a problem-solution perspective by categorizing works based on their contributions and challenges addressed. It also presents an updated view of the state-of-the-art for both application domains and provides a roadmap for future research.
Publication date: 19 Jan 2024
Project Page: https://doi.org/XXXXXXX.XXXXXXX
Paper: https://arxiv.org/pdf/2401.10185
