Robust Facial Alignment for Face Recognition

Kuan Pen Chou, Dong Lin Li, Mukesh Prasad*, Mahardhika Pratama, Sheng Yao Su, Haiyan Lu, Chin Teng Lin, Wen-Chieh Lin

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Scopus citations


This paper proposes a robust real-time face recognition system that utilizes regression tree based method to locate the facial feature points. The proposed system finds the face region which is suitable to perform the recognition task by geometrically analyses of the facial expression of the target face image. In real-world facial recognition systems, the face is often cropped based on the face detection techniques. The misalignment is inevitably occurred due to facial pose, noise, occlusion, and so on. However misalignment affects the recognition rate due to sensitive nature of the face classifier. The performance of the proposed approach is evaluated with four benchmark databases. The experiment results show the robustness of the proposed approach with significant improvement in the facial recognition system on the various size and resolution of given face images.

Original languageEnglish
Title of host publicationNeural Information Processing - 24th International Conference, ICONIP 2017, Proceedings
EditorsDerong Liu, Shengli Xie, El-Sayed M. El-Alfy, Dongbin Zhao, Yuanqing Li
PublisherSpringer Verlag
Number of pages8
ISBN (Print)9783319700892
StatePublished - 1 Jan 2017
Event24th International Conference on Neural Information Processing, ICONIP 2017 - Guangzhou, China
Duration: 14 Nov 201718 Nov 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10636 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference24th International Conference on Neural Information Processing, ICONIP 2017


  • Face alignment
  • Face recognition
  • Facial feature localization and sparse representation classifier


Dive into the research topics of 'Robust Facial Alignment for Face Recognition'. Together they form a unique fingerprint.

Cite this