Welcome to Hongyu Xu’s Homepage!
I am a final year Ph.D. candidate in the Department of Electrical and Computer Engineering at University of Maryland, College Park. My research advisor is Prof. Rama Chellappa. I am also affiliated with the University of Maryland Institute for Advanced Computer Studies (UMIACS).
I have worked on several topics such as deep learning for generic object detection, image/video face recognition/verification, domain adaptation, dictionary learning and sparse representation. My research interests include pattern recognition, deep learning, computer vision and machine learning.
I have done one internship at PARC-Xerox (Palo Alto Research Center) and two at Snap Research. I received my M.S. degree along the way from University of Maryland in 2016 and B.S. degree in Electrical Engineering and Information Science from University of Science and Technology of China in 2012.
- Apr, 2019. Officially Dr. Hongyu Xu.
- Mar, 2019. Our paper “A Fast and Accurate System for Face Detection, Identification, and Verification” accepted in IEEE Transactions on Biometrics, Behavior, and Identity Science (TBIOM).
- Feb, 2019. Our paper “Thresholded Features with Sparse Support Recovery” accepted in IEEE Transactions on Circuits and Systems for Video Technology (TCSVT). For the first time, we presented two theoretical recovery guarantees for the fast thresholded feature to exactly recover the nonzero support of the true sparse code, from which we also witnessed more hidden links between sparse coding and deep learning. Congratulations to Zhangyang (Atlas) Wang, Haichuan Yang, Ding Liu and Ji Liu.
- Nov, 2018. New preprint: Deep Regionlets: Blended Representation and Deep Learning for Generic Object Detection (Submitted to IEEE TPAMI Journal). Available on arXiv
- Sep, 2018. New preprint on face detection/recognition/verification: A Fast and Accurate System for Face Detection, Recognition and Verification (Submitted to IEEE TBIOM Journal first issue). Available on arXiv
- Jul, 2018. Our object detection paper “Deep Regionlets for Object Detection” accepted in ECCV 2018. Congratulations to Xutao Lv, Xiaoyu Wang, Zhou Ren and all co-authors.
- Apr, 2018. New preprint: Cross-Domain Visual Recognition via Domain Adaptive Dictionary Learning. Available on arXiv
- Apr, 2018. New preprint: Thresholded Features with Sparse Support Recovery. Available on arXiv
- Apr, 2018. New preprint: Crystal Loss and Quality Pooling for Unconstrained Face Verification and Recognition. Available on arXiv
- Mar, 2018. New preprint: Deep Regionlets for Object Detection. Available on arXiv
- Sep - Dec, 2017. Research Intern at Snap Research. Research on generic object detection using deep learning techniques.
- May - Aug, 2017. Research Intern at Snap Research, Venice, California. I worked with Dr. Xutao Lv, Dr. Xiaoyu Wang and Dr. Zhou Ren on generic object detection using deep learning techniques.