Zhixiang Min

I am a Ph.D. student (2018-) in Computer Science at Stevens Institute of Technology, advised by Professor Enrique Dunn.
I spent summer internships at Zillow and NEC Labs.
My research interest lies in 3D computer vision (SfM, SLAM, localization, reconstruction) and recently image-based rendering. My works during PhD focus on closing the gap between deep learning methods and 3D vision problems. Specifically, I design localization algorithms that ultilizes deep learning to improve challenging instances and solve ill-posed problems, while, importantly, maintaining the geometric interpretabilities.
Email: zmin1 [at] stevens.edu
[Google Scholar] / [LinkedIn] / [CV]
Projects
(I occasionally fix some typos in our camready submissions. Please refer to the arxiv version for best experience.)
![]() |
NeurOCS: Neural NOCS Supervision for Monocular 3D Object Localization
Zhixiang Min, Bingbing Zhuang, Samuel Schulter, Buyu Liu, Enrique Dunn, Manmohan Chandraker CVPR 2023 [arxiv] [appendix] [video] [cvpr_poster_session] |
![]() |
LASER: LAtent SpacE Rendering for 2D Visual Localization
Zhixiang Min, Naji Khosravan, Zachary Bessinger, Manjunath Narayana, Sing Bing Kang, Enrique Dunn, Ivaylo Boyadzhiev CVPR 2022 (Oral Presentation) [arxiv] [appendix] [video] [code] |
![]() |
VOLDOR-SLAM: For the times when feature-based or direct methods are not good enough
Zhixiang Min, Enrique Dunn ICRA 2021 [arxiv] [video] [code] Our SLAM solution has achieved 2nd place in both tracks (monocular, stereo) of CVPR'20 SLAM Challenge while retaining realtime. |
![]() |
VOLDOR: Visual Odometry from Log-logistic Dense Optical flow Residuals
Zhixiang Min, Yiding Yang, Enrique Dunn CVPR 2020 (Oral Presentation) [arxiv] [appendix] [video] [code] |