Quatnet: Quaternion-based head pose estimation with multiregression loss

Heng Wei Hsu*, Tung Yu Wu, Sheng Wan, Wing Hung Wong, Chen-Yi Lee

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

120 Scopus citations

Abstract

Head pose estimation has attracted immense research interest recently, as its inherent information significantly improves the performance of face-related applications such as face alignment and face recognition. In this paper, we conduct an in-depth study of head pose estimation and present a multiregression loss function, an L2 regression loss combined with an ordinal regression loss, to train a convolutional neural network (CNN) that is dedicated to estimating head poses from RGB images without depth information. The ordinal regression loss is utilized to address the nonstationary property observed as the facial features change with respect to different head pose angles and learn robust features. The L2 regression loss leverages these features to provide precise angle predictions for input images. To avoid the ambiguity problem in the commonly used Euler angle representation, we further formulate the head pose estimation problem in quaternions. Our quaternion-based multiregression loss method achieves state-of-The-Art performance on the AFLW2000, AFLW test set, and AFW datasets and is closing the gap with methods that utilize depth information on the BIWI dataset.

Original languageEnglish
Article number8444061
Pages (from-to)1035-1046
Number of pages12
JournalIEEE Transactions on Multimedia
Volume21
Issue number4
DOIs
StatePublished - Apr 2019

Keywords

  • Convolutional neural network (CNN)
  • head pose estimation
  • ordinal regression
  • quaternion

Fingerprint

Dive into the research topics of 'Quatnet: Quaternion-based head pose estimation with multiregression loss'. Together they form a unique fingerprint.

Cite this