Zhangyang (Atlas) Wang​​

​​​

• Tenure-Track Assistant Professor, CSE@TAMU, 2017- present.
  
• Leading the Visual Informatics@Texas A&M (VITA) group - Check Ph.D. opportunities to work with us!
• Ph.D., ECE@UIUC (2016). B.E., EEIS@USTC (2012). 
  
• Highly motivated researcher, in machine learning, deep learning, computer vision, image processing, and optimization.
  
• For up-to-date research topics that we are working on: please click "Resume", then check under "Research Interests".
  
• Co-authored 5+ journal papers, 40+ conference & workshop papers, and 1 book [Google Scholar]. Co-held 3 granted patents.
  
• Internship with Microsoft Research Redmond (2015), Adobe Research (2014), and US Army Research Lab (2013).
  
• Received 20+ scholarships and awards.
  
• Research covered by BBC, Fortune, IBT, TAMU and UIUC news, etc.​​

• [Publicity & Service] 09/2018, I was appointed as an Area Chair for ICME'19.
• [Travel & Talk] 09/2018, I gave three invited talks, at Artificial Intelligence in Medicine Seminar Series, University of Texas Southwestern Medical Center (UTSW); CS@Stevens Institute of Technology; and TAMU ECE ISS Seminar. 
  
• [Publication]  09/2018, Two papers accepted to NIPS'18. One as spotlight oral and the other as poster. Both papers and associated codes will come soon. Congratulations to my students Xiaohan and Nitin on publishing their first works in VITA! 
  
  • The spotlight oral paper "Theoretical Linear Convergence of Unfolded ISTA and its Practical Weights and Thresholds" for the first time rigorously established that the unfolded ISTA (known as LISTA for sparse signal recovery) could achieve asymptotic linear convergence (as upper bound). It also introduced practical weight structures and support detection techniques to facilitate the faster convergence. 
    
  • The poster paper "Can We Gain More from Orthogonality Regularizations in Training Deep Networks?"   developed and evaluated a number of orthogonality regularizations for training deep CNNs. It showed with rigorous experiments on state-of-the-art models that orthogonality can improve test performance as well as stabilize the training process.
• [Publication]  09/2018, One paper accepted to ICDM'18 workshop on developmental learning (DELL).  Congratulations to my student Jianghao on publishing his first work in VITA! 
  
• [Publicity & Service]  08/2018, I serve as a Guest Editor for IEEE Access, Special Issue on "Advanced Optical Imaging for Extreme Environments". High-quality submissions are solicited until 03/01/2019. [Call For Papers]
• [Travel & Talk] 08/2018, I was invited to participate in Microsoft Research Faculty Summit 2018.
• [Honor & Support] 07/2018, the TAMU team won the 2nd place in ECCV 2018 ChaLearn Challenge Track 3 (Fingerprint Denoising and Inpainting)! Congratulations to our teammates Ramakrishna Prabhu, Xiaojing Yu, Ding Liu, and Dr. Anxiao (Andrew) Jiang. [ECCV Workshop Paper] [Codes Available]
• [Travel & Talk] 07/2018, I attended ICML'18 and IJCAI'18, in Stockholm, Sweden, to present two conference papers and one workshop paper from our group, as well as to chair one conference session and one workshop.
• [Publication]  07/2018, One technical demo accepted to ACM MM'18. That was extended from a student course project in the CSCE 633 that I taught in Spring 2018.
• [Publication] 07/2018, One paper accepted to ECCV'18. We conduct a pilot study on privacy-preserving visual recognition (an increasingly demanded feature in smart home cameras, etc.), using novel algorithms derived from adversarial deep learning. Its short version appeared in ICML'18 PiMLAI workshop. [Codes Available] Congratulations to my student Zhenyu on publishing his first work in VITA! 
• [Publication]  07/2018, One paper accepted to IEEE TIP. And I'm proud of this one very much.
  • We released the first large-scale benchmark dataset for image dehazing: REalistic Single Image DEhazing (RESIDE), consisting of several sets of synthetic and real-world hazy images. 
  • Our paper presents a comprehensive study and evaluation using three different sets of criteria (objective, subjective, task-driven), for nine state-of-the-art dehazing algorithms. 
  • Being the first-of-its-kind efforts, extensive experiments on RESIDE shed light on the comparisons and limitations of state-of-the-art dehazing algorithms, and suggest promising future directions.​
• [Publication]  07/2018, a Preprint page is on, showcasing latest posted works.
• [Publication]  06/2018, One paper accepted to IISE Transactions on Healthcare Systems Engineering. The paper summarized the interdisciplinary work that I contributed to during my UW visit: detecting depression from daily life by jointly using methods from computer vision, speech signal processing and sentiment analysis. 
• [Travel & Talk] 06/2018, I gave an invited short course on deep learning /computer vision, plus an invited talk, at Institute of Mathematics and Statistics, University of São Paulo, Brazil. 
• [Travel & Talk] 06/2018, I gave two invited talks at Italy: i) the "Data-Driven Approaches in Imaging Science" Minisymposium, SIAM Imaging Science Conference 2018, Bologna; ii) Dept. of Information Engineering and Computer Science, University of Trento.
• [Publicity & Service]  06/2018, I will serve on Technical Program Committees (TPC) of AAAI'19, MIPR'19, ICDM'18, and ACCV'18.
• [Publicity & Service] 05/2018, I was appointed as an Area Chair for WACV'19.
• [Honor & Support] 05/2018, the TAMU-PKU team won the CVPR 2018 UG2 Prize Challenge! Kudos to my students Ye Yuan and Junru Wu for leading this great effort, and to our amazing friends in Peking University (Yueyu Hu, Wenhan Yang, Xiaoshuai Zhang, and Dr. Jiaying Liu). [TAMU News] [Codes Available] 
• [Publication] 05/2018, One paper accepted to ICML'18. We proposed a simple yet effective scheme for compressing very deep CNNs (GoogleNet, Wide ResNet, etc.) though re-training with a clustering regularization inspired by k-means. We additionally propose an improved set of metrics to estimate energy consumption of CNN hardware implementations. [Codes Available] Congratulations to my student Junru on publishing his first work in VITA!  
• [Publicity & Service]  05/2018, I will serve on Technical Program Committees (TPC) of NIPS'18.
• [Publication]  05/2018, One paper accepted to KDD'18 [long oral]. We presented Subspace Network, an efficient deep modeling approach for non-linear multi-task censored regression. Under mild assumptions, each layer can recover the underlying parametric subspace using only one pass of training data. It outperforms state-of-the-arts in predicting neurodegenerative clinical scores from brain imaging. [Codes Available]
• [Publicity & Service] 04/2018, I helped co-organize a challenge competition on single image dehazing at China MM'18 conference, using our recently released  RESIDE dataset. [Competition Site - in Chinese]. 
• [Travel & Talk] 04/2018, I gave an invited talk, at the Machine Learning Seminar Series, Michigan State University, on robust computer vision in outdoor environments, with dehazing as an example.
• [Publication]  04/2018, One paper accepted to IJCAI'18. We presented the first-of-its-kind study on the mutual influence between image denoising and high-level computer vision tasks. [Codes Available]
• [Travel & Talk] 04/2018, I gave an invited talk, at Walmart Tech, Dallas, TX.
• [Publication] 03/2018, Two research abstracts accepted to Alzheimer's Association International Conference (AAIC 2018). Congratulations to my student Ye on publishing her first works in VITA! 
• [Publication] 03/2018, One paper accepted to IEEE TIP. 
• [Publication] 03/2018, The book "Applied Cloud Deep Semantic Recognition: Advanced Anomaly Detection" is published by CRC Press-Tylor & Francis, and is available on Amazon.com now. I have an invited chapter there: "Deep Learning for Font Recognition and Retrieval".
• [Mentoring & Teaching] 03/2018, five out of my six Ph.D. students accept summer research internship offers from prestigious research institutions: Zhenyu Wu (US Army Research Lab), Jianghao Shen (Bell Labs), Ye Yuan (Walmart Tech), Junru Wu (NEC Labs), and Sicheng Wang (University of Syndey, as a visiting researcher). So proud of their strong competency in 1st Ph.D. year.
• [Mentoring & Teaching] 03/2018, I am truly delighted to welcome 1 Ph.D. student, and 5 visiting students, to join VITA group in CSE@TAMU - glad to see our group to grow with such a momentum!
• [Mentoring & Teaching] Fall 2018, I will teach CSCE 625: Artificial Intelligence. 
• [Travel & Talk] 03/2018, I gave four invited talks, at DARPA, Arlington VA; University of Maryland, College Park, MD; US Army Research Lab, Adelphi, MD; and TAMU HRPC Computational and Data Science Seminar. 
• [Publicity & Service] 03/2018, I will co-organize the 3rd International Workshop on Biomedical infOrmatics with Optimization and Machine learning (BOOM), co-located with International Joint Conference on Artificial Intelligence (IJCAI’18), as part of the Federated AI Meeting (FAIM) at Stockholm, Sweden, in July 2018. It has been a successful serial workshop and now presents several levels of publication channels  [Call For Paper].
• [Publicity & Service] 03/2018, I will serve on Technical Program Committees (TPC) of ACM MM'18 and  MICCAI'18.
• [Travel & Talk] 02/2018, I gave an invited talk, at the Samsung Research America, Dallas, TX.
• [Travel & Talk] 02/2018, I attended AAAI'18, in New Orleans, LA, to present one oral paper, two abstracts, as well as to chair two conference sessions on machine learning & computer vision.
• [Publicity & Service] 01/2018, I will serve on Technical Program Committees (TPC) of IJCAI'18, and ICIP'18.
• [Travel & Talk] 01/2018, I gave an invited talk, at the Bell Labs, Murray Hill, NJ.
• [Publicity & Service] 12/2017, Students from our group were interviewed by The Battalion and shared their opinions on: "How Artificial Intelligence will Affect Opportunity". 
• [Publicity & Service] 12/2017, On CSE@TAMU news: "Wang presents at Lenovo AI Innovation Challenge and Computing Community Consortium". 
• [Travel & Talk] 12/2017, I gave invited talks at ECE@Rice University, and at ECE@Univeristy of Houston.
• [Publicity & Service] 12/2017, I will serve on Technical Program Committees (TPC) of IJCNN'18, and ICME'18.
• [Mentoring & Teaching]  Spring 2018, I will teach CSCE 633: Machine Learning.
• [Publication] 11/2017, One paper and Two student abstracts accepted to AAAI'18.  
  • Full paper "End-to-End United Video Dehazing and Detection" fills in the blank of video dehazing using deep learning, and extends to enhancing video object detection in the presence of haze.
  • Abstract "Visual Recognition in Very Low-Quality Settings: Delving into the Power of Pre-Training" discussed our recent progress in object recognition from low-quality image & video.
  • Abstract "Predicting Depression Severity by Multi-Modal Feature Engineering and Fusion" summarized a part of my work in UW: applying machine learning to depression prediction. 
• [Publicity & Service] 10/2017, I will serve on the Technical Program Committee (TPC) of CVPR'18.
• [Publicity & Service] 10/2017, I will co-organize the workshop: "Real-World Face and Object Recognition from Low-Quality Images (FOR-LQ)", in conjunction with IEEE FG 2018, jointly with Prof. Dong Liu, Prof. Weisheng Dong, and Ding Liu.  
• [Honor & Support] 10/2017, I'm grateful to be selected as a winner of the Lenovo AI Innovation Challenge, and will give an invited talk at the Supercomputing (SC) 2017 conference, Denver, CO. 
• [Travel & Talk] 09/2017, I'm grateful to be selected as an invited presenter at the CCC Symposium: "Computing Research: Addressing National Priorities and Societal Needs 2017", Washington, DC.
• [Travel & Talk] 09/2017, I gave several talks on campus: ECE CESG Seminar (sparsity in deep network algorithms and hardware); CSCE Graduate Seminar and ECE Bio-Seminar (robust computer vision in the wild).
• [Publicity & Service] 09/2017, I will serve on the Technical Program Committee (TPC) of PAKDD'18.
• [Honor & Support] 08/2017, I was selected to receive the ICCV 2017 Young Researcher Travel Award.
• [Publicity & Service] 08/2017, I will serve on the Technical Program Committee (TPC) of MIPR'18.
• [Mentoring & Teaching] Fall 2017, I will teach CSCE 689: Machine Learning Methods for Computer Vision. Aggies who are interested in my research: please consider taking this course.
• [Travel & Talk] 07/2017, I attended CVPR'17, in Honolulu, HI, and gave the tutorial, "Dealing with Reality: Low-Quality Visual Data Processing and Analytics", jointly with Dr. Haichao Zhang and Prof. Jiebo Luo. [Slides uploaded to the Tutorial page] [Video record available on YouTube] [Photo]
• [Publication] 07/2017, Two papers accepted to ICCV'17: 
  • The paper "Robust Video Super-Resolution with Learned Temporal Dynamics" proposed a temporal adaptive neural network for video super-resolution that adaptively modulates temporal dependency, and reduced the complexity of motion between neighboring frames using a spatial alignment network. 
  • The paper “AOD-Net: All-in-One Dehazing Network” proposed the first-of-its-kind deep end-to-end model (defined as a fully learnable deep regression from hazy images to clean images) for single-image haze removal, and discussed its inspiring impacts on object detection [Codes Available]. Congratulations to my mentee Boyi on publishing her first paper! 
• [Publication] 07/2017, One paper accepted to ACII'17 [Codes Available]. This is a cute paper (although quite preliminary) that I particularly like: discussing how deep learning could help video coding and transmission in an end-to-end and task-driven manner. Congratulations to my mentee Bowen on publishing his first paper! 
• [Travel & Talk] 07/2017, I gave an invited talk, at the Banner Alzheimer's Institute, Phoenix, AZ.
• [Publication] 06/2017, One paper accepted to BMVC'17 [Codes Available]. Two highlights: (1) we relaxed the classical "spread property" of ℓ∞-minimization [Studer et. al. 2014] for hashing, avoiding the full-spark dictionary assumption and making anti-sparse solutions attainable under milder conditions; (2) The model learned a (dimension-reduction) transform, a dictionary and the representation simultaneously.
• [Publicity & Service] 06/2017, I will serve on the Technical Program Committee (TPC) of AAAI'18.
• [Publication] 05/2017, One paper accepted to ICIP'17.   
• [Publication] 05/2017, One paper accepted to IEEE TIP.  We proposed a greedy batch-based minimum-cost flow approach for tracking multiple objects. [Codes Available].
• [Publication] 05/2017, One paper accepted to CVPR'17 NITRE workshop. The UIUC team participated in the NITRE Image Super-Resolution Challenge, and ranked Top-5 in accuracy and Top-2 in speed, among a total of 19 teams. [Codes Available] [Read Our Co-Authored Challenge Report]
• [Publicity & Service] 04/2017, I will co-chair the VCIP'17 special session: "Regularization Techniques for High-Dimensional Visual Data Processing and Analysis", jointly with Dr. Xi Peng and Dr. Sheng Li. 
• [Publication] 04/2017, One paper accepted to IJCAI'17.
• [Publicity & Service] 03/2017, I will serve as the Program Committee Co-Chair of the 7th IEEE International Workshop on Analysis and Modeling of Faces and Gestures (AMFG), in conjunction with ICCV 2017. Please consider submitting to this highly reputed serial workshop. 
• [Mentoring & Teaching]  03/2017, I am truly delighted to welcome 4 Ph.D. students to join my research group in CSE@TAMU, at Fall 2017 -- HOWDY TEAM!
• [Publicity & Service] 03/2017, I served on the Technical Program Committee (TPC) of ACM MM'17.
• [Publicity & Service] 02/2017, I was appointed as an Area Chair for ICIP'17.
• [Honor & Support] 02/2017, I was honored and humbled to be nominated as a Finalist of the UW Amazon Catalyst Innovation Award (jointly with Prof. Shuai Huang), for our proposed smart healthcare research.
• [Publication] 02/2017, One paper accepted to IJCNN'17.
• [Announcement] 01/2017 - 07/2017, I am visiting the University of Washington, Seattle, WA, as a Research Scientist. My research focus over this period will be to develop tailored computer vision and machine learning techniques for medical and healthcare data analytics.
• [Publicity & Service] 01/2017, I was appointed as an Associate Editor for the Asia-Pacific Journal of Neural Networks and Its Applications.
• [Publicity & Service]  01/2017, I serve as a Guest Editor for IEEE Transactions on Neural Networks and Learning Systems (TNNLS), Special Issue on "Discriminative Learning for Model Optimization and Statistical Inference". High-quality submissions are solicited until the deadline 07/15/2017. [Call For Papers]
• [Publicity & Service] 01/2017,  I will serve on the Technical Program Committee (TPC) of VCIP'17.
• [Publicity & Service] 12/2016, I will co-organize the 2nd International Workshop on Biomedical infOrmatics with Optimization and Machine learning (BOOM), co-located with International Joint Conference on Artificial Intelligence (IJCAI’17), Melbourne, in August 2017. [Call For Papers].
• [Publicity & Service] 12/2016, I served on the Technical Program Committee (TPC) of IJCAI'17.
• [Publication]12/2016, My first US patent was granted! "Font recognition and font similarity learning using a deep neural network", US patent no. 9501724. 

• [Publication]12/2016, My second book will be published by Elsevier: "Deep Learning through Sparse and Low-Rank Modeling" (jointly with Prof. Yun Raymond Fu and Prof. Thomas Huang).

• [Publicity & Service] 11/2016, I served on the Technical Program Committee (TPC) of CVPR'17.

• [Announcement] 10/2016, I passed my final defense and received my Ph.D. degree from ECE@UIUC. My thesis title is: "Task-Specific and Interpretable Feature Learning".

• [Announcement] Happy 80-th Birthday, Tom! 10/2016, We held a successful symposium to honor the unparallel career of my Ph.D. advisor, Prof. Thomas S. Huang. 

• [Mentoring & Teaching] 09/2016, My mentee Bowen Cheng (senior at ECE@UIUC) received the prestigious Emma L. Marshall Scholarship from the ECE department. Congratulations, Bowen!  

• [Publicity & Service] 08/2016, I served on the Technical Program Committee (TPC) of IJCNN'17.

• [Publicity & Service] 08/2016, I served on the Technical Program Committee (TPC) of PAKDD'17.

• [Publicity & Service] 08/2016, our regular tutorial, "Deep Learning Meets Model Optimization and Statistical Inference", was accepted by ECCV 2016, October 8th, 2016, Amsterdam, the Netherlands.
• [Travel & Talk] 07/2016, I was a short-term visiting researcher, at Dept. Electrical & Computer Engineering (ECE), University of Minnesota, Minneapolis, MN.
• [Publicity & Service] 07/2016, my research was featured by UIUC ECE Headline again.
• [Travel & Talk] 07/2016,  I attended IJCAI'16, in New York, NY, and presented one paper. I also co-chaired the IJCAI'16 BOOM workshop. 
• [Travel & Talk] 06/2016, I attended CVPR'16, in Las Vegas, NV, and presented two papers (1, 2). I was also selected into the CVPR 2016 Doctoral Consortium (mentor: Prof. Rene Vidal).
• [Publication] 06/2016, One paper accepted to ACM MM'16.
• [Publicity & Service] 06/2016, DeepFont was covered by BBS News and International Business Times. 
• [Travel & Talk] 06/2016, I was an invited speaker at the PKU Workshop on Data Science and Applications, held by Beijing International Center for Mathematical Research (BICMR), PKU, Beijing, China.
• [Travel & Talk] 06/2016, I was an invited speaker at the 2016 Workshop on Sparsity Revisited (SPARE), USTC, Hefei, China.
• [Travel & Talk] 05/2016, I gave an invited talk, in Chinese R Conference 2016 (Machine Learning Session), organized by Capital of Statistics (COS), Renmin University of China, Beijing, China.
• [Honor & Support] 05/2016, I felt honored to accept the Dissertation Completion Fellowship, as the only recipient from the ECE department. This year, 29 Ph.D. students from 27 departments were selected for the prestigious fellowship, by UIUC graduate college.
• [Travel & Talk]  05/2016, I attended SDM'16, in Miami, FL, and presented one paper. I also helped organize the SDM'16 MLRec workshop.
• [Travel & Talk] 05/2016, I visited Dept. EECS, Northwestern University, Chicago, IL, and gave a seminar, hosted by Prof. Ying Wu.
• [Announcement] 04/2016, I accepted the offer to join the Computer Science and Engineering (CSE) Department, at the Texas A&M University (TAMU), as a Tenure-Track Assistant Professor. 
• [Travel & Talk] 04/2016, I visited Dept. Industrial & Systems Engineering, University of Washington, Seattle, WA, and gave a seminar, hosted by Prof. Shuai Huang.
• [Honor & Support] 04/2016, I was honored and humbled to accept the 2016 Thomas and Margaret Huang Award for Graduate Research from UIUC, for my research record.
• [Publication] 04/2016, One paper accepted to IJCAI'16.
• [Publicity & Service] 04/2016, DeepFont was covered by Fortune, Adobe Corporate Communications, and UIUC Resonance Magazine (Spring 2016).
• [Publicity & Service] 04/2016, I served on Technical Program Committee (TPC) of ACM MM'16.
• [Publication] 03/2016, Two papers accepted to CVPR'16.
• [Honor & Support] 02/2016, I received the Chinese Government Award for Outstanding Graduates Abroad,  for both academic and research excellences.
• [Publicity & Service] 02/2016, I was elected as an online organization committee member of VALSE. 
• [Travel & Talk] 02/2016, I attended AAAI'16, in Phoenix, AZ, and presented two papers.
• [Honor & Support] 01/2016, I acknowledged Microsoft for their donation of three Microsoft Bands II, to support my research on deep learning applied to healthcare data analytics.
• [Publication] 01/2016, my co-authored book became available on Amazon.com: "Sparse Coding and Its Applications in Computer Vision".
• [Honor & Support] 12/2015, I received the AAAI scholarship, and was selected as a "finalist" for the AAAI best student paper presentation competition.
• [Honor & Support] 12/2015, I acknowledged NVidia for their donation of One Titan X GPU card, to support my research on deep learning applied to food dietary assessment.
• [Publication]  12/2015,  One paper accepted to SDM'16. 
• [Travel & Talk] 12/2015, I gave an invited talk to CS@UTSA through webinar, hosted by Prof. Qi Tian.
• [Honor & Support] 11/2015, I received the highly competitive and prestigious Baidu Research Fellowship, among a total of 10 global recipients.
• [Publication]  11/2015,  One paper and One student abstract accepted to AAAI'16.
• [Travel & Talk] 11/2015, I gave invited talks in MSRA, Baidu Research, and EEIS@USTC.
• [Publicity & Service]11/2015, DeepFont was extensively covered by media, including UIUC ECE Headline News, among many others.
• [Travel & Talk] 10/2015, I attended ACM MM'15, in Brisbane, Australia, to present one full (oral) paper and one technical demo, both about DeepFont.
• New Website In Use .......