Concordia University

http://www.concordia.ca/content/shared/en/news/encs/computer-science/2017/11/09/Doctoral-Thesis-Defense-Chi-Nhan-Duong.html

notice

Doctoral Thesis Defense: Chi Nhan Duong

November 9, 2017

Speaker: Chi Nhan Duong

Supervisors: Drs. T. D. Bui, K. Luu

Supervisory Committee: Drs. A. Krzyzak, T. Popa, H. Rivaz, B. V. K. Vijaykumar, H. Ge (Chair)

Title: Beyond PCA: Deep Learning Approaches for Face Modeling and Aging

Date: Thursday, November 9, 2017

Time: 14:00

Place: EV 3.309

ABSTRACT

Modeling faces with large variations has been a challenging task in computer vision. These variations such as expressions, poses and occlusions are usually complex and non-linear. Moreover, new facial images also come with their own characteristic artifacts that greatly diverse. Therefore, a good face modeling approach needs to be carefully designed for flexibly adapting to these challenging issues. In this thesis, we present two novel approaches, i.e. Deep Appearance Models (DAM) and Robust Deep Appearance Models (RDAM), to accurately capture both shape and texture of face images under large variations. In DAM, three crucial components represented in hierarchical layers are modeled using Deep Boltzmann Machines to robustly capture the variations of facial shapes and appearances. Then RDAM, an improved version of DAM, is introduced to better handle the occluded face areas and, therefore, produces more plausible results. These approaches are evaluated in various applications to demonstrate their robustness and capabilities in handling occlusions, facial representation, and reconstruction using challenging face databases.

In addition to DAM and RDAM that are mainly used for modeling single facial image, the second part of the thesis focuses on novel deep models, i.e. Temporal Restricted Boltzmann Machines (TRBM) and tractable Temporal Non-volume Preserving (TNVP) approaches, to further model face sequences. In the applications of face age progression, age regression, and age-invariant face recognition, these models have shown their potential not only in efficiently capturing the non-linear age-related variance but also

producing a smooth synthesis in age progression across faces. Moreover, the structure of TNVP can be transformed into a deep convolutional network while keeping the advantages of probabilistic models with tractable log-likelihood density estimation. The proposed approaches are evaluated in both facial age progression and cross-age face verification and consistently show the state-of-the-art results in various face aging databases, i.e. FG-NET, MORPH, our collected large-scale aging database named AginG Faces in the Wild (AGFW), and Cross-Age Celebrity Dataset (CACD). A large-scale face verification on Megaface challenge 1 is also performed to further show the advantages of our proposed approaches.

 




Back to top

© Concordia University