Facial anatomical landmark detection using regularized transfer learning with application to Fetal Alcohol Syndrome recognition

Zeyu Fu*, Jianbo Jiao, Michael Suttie, J. Alison Noble

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)
79 Downloads (Pure)

Abstract

Fetal alcohol syndrome (FAS) caused by prenatal alcohol exposure can result in a series of cranio-facial anomalies, and behavioral and neurocognitive problems. Current diagnosis of FAS is typically done by identifying a set of facial characteristics, which are often obtained by manual examination. Anatomical landmark detection, which provides rich geometric information, is important to detect the presence of FAS associated facial anomalies. This imaging application is characterized by large variations in data appearance and limited availability of labeled data. Current deep learning-based heatmap regression methods designed for facial landmark detection in natural images assume availability of large datasets and are therefore not wellsuited for this application. To address this restriction, we develop a new regularized transfer learning approach that exploits the knowledge of a network learned on large facial recognition datasets. In contrast to standard transfer learning which focuses on adjusting the pre-trained weights, the proposed learning approach regularizes the model behavior. It explicitly reuses the rich visual semantics of a domain-similar source model on the target task data as an additional supervisory signal for regularizing landmark detection optimization. Specifically, we develop four regularization constraints for the proposed transfer learning, including constraining the feature outputs from classification and intermediate layers, as well as matching activation attention maps in both spatial and channel levels. Experimental evaluation on a collected clinical imaging dataset demonstrate that the proposed approach can effectively improve model generalizability under limited training samples, and is advantageous to other approaches in the literature.
Original languageEnglish
Article number9531054
Pages (from-to)1591-1601
Number of pages11
JournalIEEE Journal of Biomedical and Health Informatics
Volume26
Issue number4
Early online date8 Sept 2021
DOIs
Publication statusPublished - Apr 2022

Bibliographical note

Funding Information:
Manuscript received December 28, 2020; revised August 2, 2021; accepted September 1, 2021. Date of publication September 8, 2021; date of current version April 13, 2022. This work was done in conjunction with the Collaborative Initiative on Fetal Alcohol Spectrum Disorders (CIFASD), which was supported by grants from the National Institute on Alcohol Abuse and Alcoholism (NIAAA), in part by NIH under Grant U01AA014809, and in part by EPSRC under Grant EP/M013774/1. (Corresponding author: Zeyu Fu.) Zeyu Fu, Jianbo Jiao, and J. Alison Noble are with the Department of Engineering Science, University of Oxford, OX1 2JD Oxford, U.K. (e-mail: zeyu.fu@eng.ox.ac.uk; jianbo.jiao@eng.ox.ac.uk; alison.noble@eng.ox.ac.uk).

Publisher Copyright:
© 2013 IEEE.

Keywords

  • Face/diagnostic imaging
  • Female
  • Fetal Alcohol Spectrum Disorders/diagnostic imaging
  • Humans
  • Machine Learning
  • Pregnancy
  • Prenatal Exposure Delayed Effects
  • Semantics

ASJC Scopus subject areas

  • Computer Science Applications
  • Health Informatics
  • Electrical and Electronic Engineering
  • Health Information Management

Fingerprint

Dive into the research topics of 'Facial anatomical landmark detection using regularized transfer learning with application to Fetal Alcohol Syndrome recognition'. Together they form a unique fingerprint.

Cite this