Change search
ReferencesLink to record
Permanent link

Direct link
Human Face Identification and Face Attribute Prediction: From Gabor Filtering to Deep Learning
KTH, School of Computer Science and Communication (CSC).ORCID iD: 0000-0002-8673-0797
2016 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

After decades of research, it is exciting to see that face recognition technology has entered a most flourishing era. Driven by the latest development in data science and especially technical evolutions in computer vision and pattern recognition, face recognition has achieved significant progress over the last three years. In the near future, people can expect many useful and interesting face recognition applications to be deployed in many situations: they can be used for identifying suspects, organizing your photos with family and friends, and making computers better understand human beings. Many mysterious face recognition tricks depicted in movies may become reality in several years' time.

This thesis focuses on the development of face recognition algorithms that identify people from a single still image. Two questions are specifically studied. First, it introduces how we identify faces captured in controlled scenarios with cooperative users. In this scenario, a face recognition system captures a face and finds the most similar face from the ones stored in the face recognition system. Second, it describes our solutions for predicting face attributes from faces captured under arbitrary imaging conditions. These two problems were tackled by different schools of technologies: the solution to the first question employed a learning-free approach, whereas the latter question was solved by using the most recent Deep Learning technology. Thus, this thesis also reflects the technological evolution of face recognition over recent years.

To identify faces in controlled scenarios, we propose a novel Block Matching approach, which can effectively match faces without feature engineering or any machine learning components. By representing faces with very concise Gabor phase codes and matching them through our Block Matching approach, the identification accuracy is entirely comparable to and even better than the state-of-the-art. For predicting the attributes from faces captured in the wild, we propose leveraging the off-the-shelf mid-level representations from pre-trained convolutional neural networks. Comparative experiments show that our solution outperforms the previous state-of-the-art solution with a large margin in terms of both accuracy and efficiency. 

The approaches described in this thesis may look different from the ``mainstream''. But, together with the empirical findings, I hope they could provide some insights and update widely adopted concepts for solving related face recognition and computer vision problems.

Place, publisher, year, edition, pages
Stockholm: KTH Royal Institute of Technology, 2016. , 27 p.
Series
TRITA-CSC-A, ISSN 1653-5723
National Category
Engineering and Technology
Research subject
Media Technology
Identifiers
URN: urn:nbn:se:kth:diva-195092ISBN: 978-91-7729-156-5OAI: oai:DiVA.org:kth-195092DiVA: diva2:1044030
Public defence
2016-11-11, K1, 13:00 (English)
Opponent
Supervisors
Note

QC 20161103

Available from: 2016-11-03 Created: 2016-11-01 Last updated: 2016-11-16Bibliographically approved
List of papers
1. Is block matching an alternative tool to LBP for face recognition?
Open this publication in new window or tab >>Is block matching an alternative tool to LBP for face recognition?
2014 (English)Conference paper (Refereed)
Abstract [en]

In this paper, we introduce the Block Matching (BM) as an alternative patch-based local matching approach for solving the face recognition problem. The Block Matching enables an image patch of the probe face image to search for its best matching from displaced positions in the gallery face image. This matching strategy is very effective for handling spatial shift between two images and it is radically different from that of the widely used LBP type patch-based local matching approaches. Our evaluations on the FERET and CMU-PIE databases show that the performance of this simple method is well comparable (superior) to that of the popular LBP approach. We argue that the Block Matching could provide face recognition a new approach with more flexible algorithm architecture. One can expect that it could lead to much higher performance when combining with other feature extraction techniques, like Gabor wavelet and deep learning.

Place, publisher, year, edition, pages
IEEE conference proceedings, 2014
Keyword
block matching, face recognition, LBP, patch-based matching, Feature extraction, Image processing, Motion compensation, Algorithm architectures, Feature extraction techniques, Gabor wavelets, Local matching, New approaches, Patch based, Image matching
National Category
Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:kth:diva-174798 (URN)10.1109/ICIP.2014.7025145 (DOI)2-s2.0-84949929395 (ScopusID)9781479957514 (ISBN)
Conference
2014 IEEE International Conference on Image Processing, ICIP 2014
Note

QC 20151208

Available from: 2015-12-08 Created: 2015-10-07 Last updated: 2016-11-01Bibliographically approved
2. Leveraging Gabor Phase for Face Identification in Controlled Scenarios
Open this publication in new window or tab >>Leveraging Gabor Phase for Face Identification in Controlled Scenarios
2016 (English)In: Proceedings of the 11th Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, Science and Technology Publications,Lda , 2016, 49-58 p.Conference paper (Refereed)
Abstract [en]

Gabor features have been widely employed in solving face recognition problems in controlled scenarios. To construct discriminative face features from the complex Gabor space, the amplitude information is commonly preferred, while the other one — the phase — is not well utilized due to its spatial shift sensitivity. In this paper, we address the problem of face recognition in controlled scenarios. Our focus is on the selection of a suitable signal representation and the development of a better strategy for face feature construction. We demonstrate that through our Block Matching scheme Gabor phase information is powerful enough to improve the performance of face identification. Compared to state of the art Gabor filtering based approaches, the proposed algorithm features much lower algorithmic complexity. This is mainly due to our Block Matching enables the employment of high definition Gabor phase. Thus, a single-scale Gabor frequency band is sufficient for discrimination. Furthermore, learning process is not involved in the facial feature construction, which avoids the risk of building a database-dependent algorithm. Benchmark evaluations show that the proposed learning-free algorith outperforms state-of-the-art Gabor approaches and is even comparable to Deep Learning solutions.

Place, publisher, year, edition, pages
Science and Technology Publications,Lda, 2016
Keyword
Face Recognition, Controlled Scenario, HD Gabor Phase, Block Matching, Learning-free, Deep Learning
National Category
Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:kth:diva-189188 (URN)10.5220/0005723700490058 (DOI)
Conference
International Conference on Computer Vision Theory and Applications, Rome, Italy, 27-29 February 2016
Note

QC 20160715

Available from: 2016-06-28 Created: 2016-06-28 Last updated: 2016-11-01Bibliographically approved
3. How Good Can a Face Identifier Be Without Learning
Open this publication in new window or tab >>How Good Can a Face Identifier Be Without Learning
2016 (English)In: Lecture Notes in Computer Science, ISSN 0302-9743, E-ISSN 1611-3349Article in journal, Editorial material (Refereed) Published
Abstract [en]

Constructing discriminative features is an essential issue in developing face recognition algorithms. There are two schools in how features are constructed: hand-crafted features and learned features from data. A clear trend in the face recognition community is to use learned features to replace hand-crafted ones for face recognition, due to the superb performance achieved by learned features through Deep Learning networks. Given the negative aspects of database-dependent solutions, we consider an alternative and demonstrate that, for good generalization performance, developing face recognition algorithms by using handcrafted features is surprisingly promising when the training dataset is small or medium sized. We show how to build such a face identifier with our Block Matching method which leverages the power of the Gabor phase in face images. Although no learning process is involved, empirical results show that the performance of this “designed” identifier is comparable (superior) to state-of-the-art identifiers and even close to Deep Learning approaches.

Place, publisher, year, edition, pages
Springer, 2016
National Category
Computer Vision and Robotics (Autonomous Systems) Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:kth:diva-189190 (URN)
External cooperation:
Available from: 2016-06-28 Created: 2016-06-28 Last updated: 2016-11-01
4. Face Attribute Prediction Using Off-The-Shelf CNN Features
Open this publication in new window or tab >>Face Attribute Prediction Using Off-The-Shelf CNN Features
2016 (English)Conference paper (Refereed)
Abstract [en]

Predicting attributes from face images in the wild is a challenging computer vision problem. To automatically describe face attributes from face containing images, traditionally one needs to cascade three technical blocks — face localization, facial descriptor construction, and attribute classification — in a pipeline. As a typical classification problem, face attribute preiction has been addressed using deep learning. Current state-of-the-art performance was achieved by using two cascaded Convolutional Neural Networks (CNNs), which were specifically trained to learn face localization and attribute description. In this paper, we experiment with an alternative way of employing the power of deep representations from CNNs. Combining with conventional face localization techniques, we use off-the-shelf architectures trained for face recognition to build facial descriptors. Recognizing that the describable face attributes are diverse, our face descriptors are constructed from different levels of the CNNs for different attributes to best facilitate face attribute prediction. Experiments on two large datasets, LFWA and CelebA, show that our approach is entirely comparable to the state-of-the-art. Our findings not only demonstrate an efficient face attribute prediction approach, but also raise an important question: how to leverage the power of off-the-shelf CNN representations for novel tasks

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2016
National Category
Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:kth:diva-189187 (URN)10.1109/ICTON.2016.7550595 (DOI)2-s2.0-84985910600 (ScopusID)
Conference
ICB 2016 The 9th IAPR International Conference on Biometrics ,June 13-16, 2016. Halmstad, Sweden
Note

QC 20160629

Available from: 2016-06-28 Created: 2016-06-28 Last updated: 2016-11-23Bibliographically approved
5. Leveraging Mid-level Deep Representations for Prediction Face Attributes in the Wild
Open this publication in new window or tab >>Leveraging Mid-level Deep Representations for Prediction Face Attributes in the Wild
2016 (English)In: 2016 IEEE International Conference on Image Processing (ICIP), Institute of Electrical and Electronics Engineers (IEEE), 2016Conference paper (Refereed)
Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2016
National Category
Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:kth:diva-189195 (URN)10.1109/ICIP.2016.7532958 (DOI)
Conference
2016 IEEE International Conference on Image Processing (ICIP)
Note

QC 20160629

Available from: 2016-06-28 Created: 2016-06-28 Last updated: 2016-11-30Bibliographically approved
6. Transferring from Face Recognition to Face Attribute Prediction through Adaptive Selection of Off-the-shelf CNN Representations
Open this publication in new window or tab >>Transferring from Face Recognition to Face Attribute Prediction through Adaptive Selection of Off-the-shelf CNN Representations
(English)Manuscript (preprint) (Other academic)
National Category
Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:kth:diva-189197 (URN)
Note

QC 20160719

Available from: 2016-06-28 Created: 2016-06-28 Last updated: 2016-11-01Bibliographically approved

Open Access in DiVA

fulltext(9088 kB)3 downloads
File information
File name FULLTEXT01.pdfFile size 9088 kBChecksum SHA-512
b3bc287e0ed3e30e3ae68c1618097f85f216a053931c3e0dd50183d28db678e6340c9969fb93a0bf183f8cef11dca5b2734f8ea958e31d95cb5b22a19dc4c285
Type fulltextMimetype application/pdf

Search in DiVA

By author/editor
Zhong, Yang
By organisation
School of Computer Science and Communication (CSC)
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar
Total: 3 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 300 hits
ReferencesLink to record
Permanent link

Direct link