Qinxun Bai (白沁洵)

Research Scientist
Hikvision Research America
Santa Clara, CA

Email: qinxun.bai [at] gmail.com

Curriculum Vitae

About

I am a research scientist at Hikvision Research America and I am currently looking for interns! Please contact me if you are interested in working on frontier machine learning and computer vision problems. Before joining Hikvision, I worked at Vicarious for one year. I got my PhD in Computer Science at Boston University, where I was advised by Prof. Stan Sclaroff and Prof. Steven Rosenberg, and also worked with Prof. Henry Lam. Prior to BU, I obtained my Bachelor's degree at Tsinghua University and Master's degree at National Laboratory of Pattern Recognition.

I am interested in bringing new mathematical perspectives and tools to characterize essential yet underexplored aspects of machine learning problems, especially those arising in computer vision. Such efforts may lead to new principled models and algorithms suitable for solving real-world problems. I use methodologies mainly from statistics, stochastic optimization, functional analysis, and differential geometry.

Research

Qinxun Bai, Steven Rosenberg, Zheng Wu, and Stan Sclaroff
Differential Geometric Regularization for Supervised Learning of Classifiers
International Conference on Machine Learning (ICML), 2016
[Paper] [Supp] [Poster] [Project Page] [Slides]

Qinxun Bai, Henry Lam, and Stan Sclaroff
A Bayesian Approach for Online Classifier Ensemble
Under review, 2015 [arXiv] [Project Page] [Slides]

Qinxun Bai, Henry Lam, and Stan Sclaroff
A Bayesian Framework for Online Classifier Ensemble
International Conference on Machine Learning (ICML), 2014 [Paper] [Supp]

Qinxun Bai, Zheng Wu, Stan Sclaroff, Margrit Betke, and Camille Monnier
Randomized Ensemble Tracking
IEEE International Conference on Computer Vision (ICCV), 2013
[Paper] [Project Page] [Slides]

Qinxun Bai, Yihong Wu, and Lixin Fan
PCA-based Structure Refinement for Reconstruction of Urban Scene
IEEE International Conference on Image Processing (ICIP), 2010
[Paper] [Project Page]

Zhen Lei, Qinxun Bai, Ran He, and Stan Z. Li
Face Shape Recovery from a Single Image Using CCA Mapping between Tensor Spaces
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2008
[Paper] [Project Page]

Skills

  • Proficiency in C/C++ and Matlab with 10 years of experience
  • Familiar with Lua and Torch
  • CUDA C programming and layer module design experience for deep learning

Teaching