國立虎尾科技大學 |

Representation Learning and Image Synthesis for Deep Face Recognition.

Record Type:	Language materials, manuscript : Monograph/item
Title/Author:	Representation Learning and Image Synthesis for Deep Face Recognition./
Author:	Yin, Xi.
Description:	1 online resource (124 pages)
Notes:	Source: Dissertation Abstracts International, Volume: 79-12(E), Section: B.
Contained By:	Dissertation Abstracts International79-12B(E).
Subject:	Computer science. -
Online resource:	click for full text (PQDT)
ISBN:	9780438289192

Representation Learning and Image Synthesis for Deep Face Recognition.
Yin, Xi.

Representation Learning and Image Synthesis for Deep Face Recognition. - 1 online resource (124 pages)

Source: Dissertation Abstracts International, Volume: 79-12(E), Section: B.

Thesis (Ph.D.)--Michigan State University, 2018.

Includes bibliographical references

Face recognition has been advanced a lot in recent years thanks to the development of deep neural networks. The large intra-class variations in pose, illumination, and expression are the long-standing challenges. Learning a discriminative representation that is robust to these variations is the key. In the scenarios of profile pose or long-tail training data, image or feature-level data augmentation is needed. This dissertation presents three different methods to solve these problems. First, we explore a multi-task Convolutional Neural Network (CNN) that aims to leverage side tasks to improve representation learning. A pose-directed multi-task CNN is introduced to better handle pose variation. The proposed framework is effective in pose-invariant face recognition. Second, we propose a Face Frontalization-Generative Adversarial Network (FF-GAN) that can generate a frontal face even from an input image with extreme profile pose. FF-GAN handles pose variation from the perspective of image-level data augmentation. Multiple loss functions are proposed to achieve large-pose face frontalization. The proposed approach is evaluated on various tasks including face reconstruction, landmark detection, face frontalization, and face recognition. Third, a feature transfer learning method is presented to solve the problem of insufficient intra-class variation via feature-level data augmentation. A Gaussian prior is assumed across all the regular classes and the variance are transferred from regular classes to long-tail classes. Further, an alternating training regimen is proposed to simultaneously achieve less biased decision boundaries and more discriminative representations. Extensive experiments have demonstrated the effectiveness of the proposed feature transfer framework.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2018

Mode of access: World Wide Web

ISBN: 9780438289192Subjects--Topical Terms:

573171
Computer science.
Index Terms--Genre/Form:

554714
Electronic books.

Representation Learning and Image Synthesis for Deep Face Recognition.
LDR:03075ntm a2200349Ki 4500 001 918434
005 20181114145237.5
006 m o u
007 cr mn||||a|a||
008 190606s2018 xx obm 000 0 eng d
020 $a 9780438289192
035 $a (MiAaPQ)AAI10930064
035 $a (MiAaPQ)grad.msu:16300
035 $a AAI10930064
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Yin, Xi. $3 1192757
245 1 0 $a Representation Learning and Image Synthesis for Deep Face Recognition.
264 0 $c 2018
300 $a 1 online resource (124 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertation Abstracts International, Volume: 79-12(E), Section: B.
500 $a Includes supplementary digital materials.
500 $a Adviser: Xiaoming Liu.
502 $a Thesis (Ph.D.)--Michigan State University, 2018.
504 $a Includes bibliographical references
520 $a Face recognition has been advanced a lot in recent years thanks to the development of deep neural networks. The large intra-class variations in pose, illumination, and expression are the long-standing challenges. Learning a discriminative representation that is robust to these variations is the key. In the scenarios of profile pose or long-tail training data, image or feature-level data augmentation is needed. This dissertation presents three different methods to solve these problems. First, we explore a multi-task Convolutional Neural Network (CNN) that aims to leverage side tasks to improve representation learning. A pose-directed multi-task CNN is introduced to better handle pose variation. The proposed framework is effective in pose-invariant face recognition. Second, we propose a Face Frontalization-Generative Adversarial Network (FF-GAN) that can generate a frontal face even from an input image with extreme profile pose. FF-GAN handles pose variation from the perspective of image-level data augmentation. Multiple loss functions are proposed to achieve large-pose face frontalization. The proposed approach is evaluated on various tasks including face reconstruction, landmark detection, face frontalization, and face recognition. Third, a feature transfer learning method is presented to solve the problem of insufficient intra-class variation via feature-level data augmentation. A Gaussian prior is assumed across all the regular classes and the variance are transferred from regular classes to long-tail classes. Further, an alternating training regimen is proposed to simultaneously achieve less biased decision boundaries and more discriminative representations. Extensive experiments have demonstrated the effectiveness of the proposed feature transfer framework.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2018
538 $a Mode of access: World Wide Web
650 4 $a Computer science. $3 573171
650 4 $a Artificial intelligence. $3 559380
655 7 $a Electronic books. $2 local $3 554714
690 $a 0984
690 $a 0800
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a Michigan State University. $b Computer Science. $3 845509
773 0 $t Dissertation Abstracts International $g 79-12B(E).
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10930064 $z click for full text (PQDT)