The 4th RePSS - Multimodal Fusion Learning for Remote Physiological Signal Sensing

To be held at IJCAI 2025, August 2025, Montreal, Canada

Challenge

Challenge tasks are all related to the main theme of multimodal fusion learning for remote physiological signal sensing. Most current rPPG research used RGB videos, which are prone to the impacts of environmental lighting variations or skin tones. Given that NIR videos are more robust to lighting virations, and NIR facial videos also contain physiological signal information, this task aims to draw attention to innovative fusion strategies of RGB and NIR that can enhance the accuracy and reliablity of rPPG measurement especially on under challenging conditions.

This challenge will be organized on the Kaggle website. On the Kaggle website, instructions and data will be shared, and results from participants will be submitted and ranked. The top 3 teams will be awarded certificates if the top 3 teams submit papers and are present at the workshop.

Please note that researchers who 1) previously worked with the RePSS organizing team, or 2) have collaborated/are collaborating with RePSS organizers that have access to the full datasets of OBF or VIPL-HR-V2, can participate in the challenge. Their performance will be shown in the leaderboard on the Kaggle website, but won’t be counted for the final ranking and Top-3 certificate for fairness consideration. We will clarify this issue with concerned teams during team registration via email. Thank you for your understanding!


Challenge timeline