I’m a research scientist at the Nvidia Research AI/ML team. Before joining Nvidia, I received a Ph.D. from Stanford University. I had a great pleasure working with great minds at Stanford on navigation, 2D feature learning, 2D scene graph, 3D perception, 3D reconstruction, building 3D datasets, and 4D perception. Before joining the lab, I also briefly worked on convex optimization, information theory, and graphics and the experience helped me shape my research. Also, I am interested in unsupervised learning, ML algorithms, robotics, and reinforcement learning. Feel free to send an email to chrischoy@ai.stanford.edu or cchoy@nvidia.com if you have a question regarding my research.

News

Publications

Ph.D. Thesis

Preprints

Patents

  • Universal correspondence network, US Patent App. 10/115,032
  • Systems and Methods for Semantic Segmentation of 3D Point Clouds, US Patent App. 16/155,843

Open-source Projects

  • Minkowski Engine, an auto-diff library for sparse tensors code
  • Fully Convolutional Geometric Features, fast and accurate 3D feature code
  • 3D Recurrent Reconstruction Neural Network code
  • Fully Differentiable Deep Neural Decision Forest code
  • Python GPU Accelerated K-Nearest Neighbor code

Talks

Click to expand
  • Optimization in Neural Networks, Stanford Deep Learning Reading Group, CA, USA, July 14th 2017, slides.
  • Toyota Research Institute, Weakly Supervised Generative Adversarial Network for 3D Reconstruction, CA, USA, June 14th 2017.
  • Stanford Faculty Lunch, Learning to Perceive the 3D World, April 18th 2017, CA, USA, slides.
  • Apple Deep Learning Reading Group, Weakly Supervised Generative Adversarial Network for 3D Reconstruction, CA, USA, Feb 26th 2017, slides.
  • Stanford Vision Reading Group, Wasserstein GAN and Bidirectional GAN, CA, USA, January 2017, slides.
  • Neural Information Processing Systems, Oral, Universal Correspondence Networks, Barcelona, Spain, December 2016, slides.
  • European Conference on Computer Vision, Amsterdam, Netherlands, ObjectNet3D, Spotlight Oral, October 2016.
  • Korea Advanced Institute of Science and Technology, Universal Correspondence Networks and 3D Recurrent Reconstruction Neural Networks, Daejeon, Korea, June 2016.
  • NVIDIA GTC Hangout: Deep Learning in Image and Video 2016, 3D Recurrent Reconstruction Neural Networks, CA, USA, April 6th 2016.
  • Stanford Deep Learning Reading Group, Decision Forest and Deep Neural Decision Forest, CA, USA, Feb 2016, slides.

Teaching

  • CS231A Teaching Assistant, 2015

Coursework

Click to expand
  • Computer Vision
  • Machine Learning
  • Information Theory
  • Stochastic Control
  • Stochastic Processes
  • Mathematical Analysis
  • Convex Optimization I
  • Convex Optimization II
  • Mining Massive Data Sets
  • Statistical Signal Processing
  • Interactive Computer Graphics
  • Probabilistic Graphical Models
  • Modern Applied Statistics: Learning
  • Modern Applied Statistics: Data Mining
  • Convolutional Neural Networks For Visual Recognition

Education

  • Ph.D. in EE at Stanford University
  • M.S. in EE at Stanford University
  • B.S. in EE at Korea Advanced Institute of Science and Technology
    • Summa Cum Laude

Honors and Awards

  • Winner of the 2019 ScanNet Semantic Segmentation Challenge
  • Stanford SystemX FMA Fellowship
  • The Korea Foundation for Advanced Studied, Ph.D. Fellowship
  • The 5th Presidential Science Fellowship for Undergraduate Study
  • Korea Advanced Institute of Science and Technology EECS Merit Scholarship for Best Performance
  • Korea Advanced Institute of Science and Technology Alumni Chairman’s Award
  • Korea Science Olympiad, Junior High, Gold Medal