The 2nd Challenge on Remote Physiological Signal Sensing (RePSS)

Remote measurement of physiological signals from videos is an emerging topic. The topic draws great interest, but the lack of publicly available benchmark databases and a fair validation platform are hindering its further development. The RePSS Challenge is organized as an annual event for this concern. Here the 2nd RePSS is organized in conjunction with ICCV 2021. The 2nd RePSS contains two competition tracks. Track 1 is to measure inter-beat-intervals (IBI) from facial videos, which requires accurate measurement of each individual pulse peak. Track 2 is about respiration measurement from facial videos, as respiration is another important physiological index related to both health and emotional status. One new dataset is built and shared for Track 2. This paper presents an overview of the challenge, including data, protocol, results, and discussion. We highlighted the top-ranked solutions to provide insight for researchers, and we also outline future directions for this topic and this challenge.

Li Xiaobai, Sun Haomiao, Sun Zhaodong, Han Hu, Dantcheva Antitza, Shan Shiguang, Zhao Guoying

B3 Article in conference proceedings

2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW), 11-17 Oct. 2021, Montreal, BC, Canada

X. Li et al., "The 2nd Challenge on Remote Physiological Signal Sensing (RePSS)," 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW), 2021, pp. 2404-2413, doi: 10.1109/ICCVW54120.2021.00273

https://doi.org/10.1109/ICCVW54120.2021.00273 http://urn.fi/urn:nbn:fi-fe202201041144