eduzhai > Applied Sciences > Engineering >

Meta-rPPG Remote Heart Rate Estimation Using a Transductive Meta-Learner

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 26 pages

Abstract: Remote heart rate estimation is the measurement of heart rate without anyphysical contact with the subject and is accomplished using remotephotoplethysmography (rPPG) in this work. rPPG signals are usually collectedusing a video camera with a limitation of being sensitive to multiplecontributing factors, e.g. variation in skin tone, lighting condition andfacial structure. End-to-end supervised learning approach performs well whentraining data is abundant, covering a distribution that doesn t deviate toomuch from the distribution of testing data or during deployment. To cope withthe unforeseeable distributional changes during deployment, we propose atransductive meta-learner that takes unlabeled samples during testing(deployment) for a self-supervised weight adjustment (also known astransductive inference), providing fast adaptation to the distributionalchanges. Using this approach, we achieve state-of-the-art performance onMAHNOB-HCI and UBFC-rPPG.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×