Anh T Tran

  Anh T Tran, 

AI Research Scientist
VinAI Research
  v.anhtt152 at vinai.io
anh.tt at vinuni.edu.vn

I am an AI Research Scientist in VinAI Research and an affiliated faculty at Vin University. Before that, I was a Software Developer Engineer and an Applied Scientist at Amazon AWS Rekognition. My research interest is Computer Vision, and my expertise is Facial Image Processing.

I took a PhD degree from Computer Science Department of University of Southern California, under the support of the Vietnam Education Foundation funded by the U.S. government. My advisor was Prof. Gerard Medioni.

Previously, I received a BSc in Computer Science from Hanoi University of Science and Technology.

I received the first prizes in Vietnam Talent Award 2010 and Imagine Cup Vietnam 2009.

Curriculum Vitae

You can download a PDF version here


2020–now: Backdoor Attacks on Deep Neural Networks
 Research on backdoor-attack-related topics, including (1) designing novel attack mechanisms, (2) designing effective backdoor defenses, (3) helpful applications of backdoor attacks in improving/protecting deep learning systems.

  
2020–now: Image Deblurring
Research on deep-learning-based image deblurring in-the-wild.

  
2021–now: Unsupervised Category-based 3D modeling
Research on unsupervised 3D modeling on single image input, using category-specific training images.

  
2019–now: Face-related sub-topics
 Various studies on different face-related problems such as: face anti-spoofing, facial makeup transfer, face synthesis, and more.

  
2015–2018: Robust 2D face recognition with augmented deep Neural Network
 Develop an effective solution to solve face recognition in-the-wild using a very deep neural network. The key technique is to enrich training data with different augmentation (head pose, shape, expression).

  
2013–2017: Robust 3-D Face Modeling from single/multiple images
 Develop an automatic framework of reconstructing high quality 3D face (shape + texture) from in-the-wild images, which can be used in face recognition and 3D animation purpose.

  
2012–2013: 3-D Face Tracking and Modeling from a single webcam
Develop a robust, real-time 3D face tracking system providing accurate 3D head pose in the presence of large out-of-plane motions, expressions changes, and partial occlusions.


   2021:
  • P. Tran, A. Tran, K. Luu, and M. H. Nguyen. "Explore Image Deblurring via Blur Kernel Space". In CVPR 2021 [arxiv][code]
  • T. Nguyen, A. Tran, Q. Phung, and M. H. Nguyen. "Lipstick Ain't Enough: Beyond Color Matching for In-the-Wild Makeup Transfer". In CVPR 2021 [project][arxiv][code][video]
  • C. Huynh, A. Tran, K. Luu, and M. H. Nguyen. "Progressive Semantic Segmentation". In CVPR 2021 [arxiv][code][video]
  • A. Nguyen and A. Tran. "WaNet - Imperceptible Warping-based Backdoor Attack". In ICLR 2021 [link][arxiv][code]
   2020:
  • A. Nguyen and A. Tran. "Input-Aware Dynamic Backdoor Attack". In NeurIPS 2020 [link][arxiv][code]
   2019:
  • A. Tran, C. Nguyen, and T. Hassner. "Transferability and Hardness of Supervised Classification Tasks". In ICCV 2019 [pdf]
  • F. Chang, A. Tran, T. Hassner, I. Masi, R. Nevatia, and G. Medioni. "Deep, Landmark-Free FAME: Face Alignment, Modeling, and Expression Estimation". In IJCV 2019 [project] [pdf]
  2017:
  • A. Tran, I. Masi, T. Hassner, E. Paz, Y. Nirkin, and G. Medioni. "Extreme 3D Face Reconstruction: Looking Past Occlusions". In CVPR 2018 [arxiv] [code]
  • F. Chang, A. Tran, T. Hassner, I. Masi, R. Nevatia, and G. Medioni. "FacePoseNet: Making a Case for Landmark-Free Face Alignment". In FG 2017. [pdf] [arxiv] [code]
 2016:
  • A. Tran, I. Masi, T. Hassner, and G. Medioni. "Regressing Robust and Discriminative 3D Morphable Models with a very Deep Neural Network". In CVPR 2017. [project] [arxiv] [code] [bib]
  • I. Masi, A. Tran, T. Hassner, J. T. Leksut, and G. Medioni. "Do We Really Need to Collect Millions of Faces for Effective Face Recognition?" In ECCV, 2016. [project] [pdf] [code] [bib]
 2014:

  • J. Choi, A. Tran, Y. Dumortier, and G. Medioni. "Real-Time 3-D Face Tracking and Modeling Framework for Mid-Res Cam". In WACV, 2014. [pdf]



2012–2017: PhD Student - University of Southern California (USA)
Viterbi School of Engineering, Computer Science Department
2005–2010: Bachelor of Science - Hanoi University of Science & Technology (Vietnam)
School of Information and Communication Technology
  • Thesis: Map Society on Mobile                                                                 Score: 9.9/10.0
    Content: social networking which supports sharing position & place on mobile (using J2ME).
  • GPA: 8.85/10.0 (original) or 3.75/4.00 according to VEF's conversion table.
  • Classification: Very good.
  • Rank: 1/302
2002–2005: High school for gifted students, Hanoi University of Science (Vietnam)
Mathematics-specified class
  • Classification: Very good.

Work: 

2019-now: VinAI Research (Vietnam)
AI Research Scientist
  • Aug 2019 - now: Face Research
2018–2019: Amazon.com, Inc (USA)
Applied Scientist
  • Jan 2018 - Jun 2018: Work on FireTV Search
  • Jul 2018 - Jul 2019: Work on Face Research, AWS Rekognition
2012–2017: University of Southern California (USA)
PhD Student
  • Aug 2015 - Dec 2017: Challenge in-the-wild 2D face recognition problem with augmented deep Neural Network
  • Jun 2013 - Dec 2017: Robust 3-D Face Modeling from single/multiple images
  • Aug 2012 - Jun 2013: 3-D Face Tracking from a webcam
2009–2012: MIMAS team (Vietnam)
Team member - Researcher & Developer
  • May 2011 - Dec 2011: Logo Recognition Project – Creating service to recognize a logo from an image captured on mobile.
  • Jan 2009 – Jun 2010: Digital Intelligence System – Create a system to recognize human specific thought or some basic states (as blink, wink, asleep) based on EEG signal and apply it to support human life (Ex: remote control devices).
2011–2012: Emotiv Systems (Australian company creating Brain Headset and its supporting Framework to analyze brain waves)
Developer, Vietnam Team
  • Oct 2011 - Jul 2012: Emotiv Store - Creating an application to help customer find, install and manage Emotiv applications.
  • Jun 2011 - Oct 2011 & Feb 2012 - Jul 2012: port Emotiv Framework to MAC OS X.
2010–2011: Novellus Vietnam LTD, an office of Novellus Systems Inc in Vietnam
SAP Developer, ITD Department
  • Jul 2010 - Jun 2011: Modify programs and reports based on internal users’ requirements in Novellus SAP system.
    Use ABAP programming


OS Windows, Linux, Mac OSX
Database SQL Server, MySQL, Access
Language Python, Matlab, C/C++, Java, C#, ASP, PHP, UNIX shell, Assembly ...
Technology
 Deep learning with PyTorch, MXNet, Caffe, CUDA, J2EE, J2ME, .NET ...
Others
 Analyze & Design, Microsoft Office



2010 Vietnam Talent Award; Dan Tri E-newspaper, VTV and VNPT Group.
2010 Title of First-ranked Student; Faculty of Information Technology.
2010 The first prize in University-level Student Research Contest; Hanoi University of Technology, Vietnam.
2009 The first prize of Imagine Cup Vietnam; Microsoft Vietnam.
2009 Lawrence S. Ting Scholarship; Lawrence S. Ting Memorial Fund, Vietnam.
2008 Lawrence S. Ting Scholarship; Lawrence S. Ting Memorial Fund, Vietnam.
2007 Fujitsu Scholarship; Fujitsu Vietnam LTD, Vietnam.