51猎奇入口

Ke Li

I am an Assistant Professor at 51猎奇入口 in beautiful Vancouver, Canada, and a Canada CIFAR AI Chair at Amii. I was previously at Google and the Institute for Advanced Study (IAS) in Princeton, and received my Ph.D. from UC Berkeley, where I was advised by , and my bachelor's in computer science from the University of Toronto. I serve as a general chair for the . My research interests are in machine learning, computer vision and algorithms. I can be reached by e-mail at keli [at] sfu [dot] ca. While at the IAS, I organized the with - check out past seminars and on .

|

Prospective MSc/PhD Students: I will be taking on a few new students this year. If you are interested in working with me, please fill out . Due to the volume of emails I receive, I am unfortunately unable to respond to every email; however, I review submissions through the form regularly and will reach out to selected students.
Prospective 51猎奇入口Undergraduate/MPCS Students: If you are interested in working on a research or capstone project on AI or related areas, please fill out .

For a quick introduction to my research, see the following talk videos:

IAS Workshop on Theory of Deep Learning () (): this is on generative modelling and nearest neighbour search and is aimed at machine learning researchers
CMU ML/Duolingo Seminar () (): this is an extended version of the above (with more details on nearest neighbour search) and is aimed at machine learning graduate students
CIFAR Deep Learning and Reinforcement Learning Summer School () (): this is on generative modelling and is aimed at a broader audience in the style of a tutorial
IAS Special Year Seminar (): this is on meta-learning and is aimed at machine learning researchers

Research Directions

I am interested in tackling fundamental problems that cannot be solved using a straightforward application of conventional techniques. Below are the major areas that I contributed to:

Generative Modelling () (): Most generative models are latent variable models, including variational autoencoders (VAEs), generative adversarial nets (GANs) and diffusion probabilistic models. The gold standard for training generative models is with maximum likelihood estimation (MLE) — however, it is not feasible to use MLE for modern, highly expressive generative models because the marginal log-likelihood is intractable. As a result, the evidence lower bound (ELBO), a lower bound on the marginal log-likelihood, is often maximized instead. The ELBO is only a good approximation to the marginal log-likelihood if the variational distribution is close to the true posterior, and so an expressive variational distribution is required. Diffusion probabilistic models increase the expressivity of the variational distribution by taking it to be the result of applying many small transformations to an analytical distribution, but do so at the expense of sampling time. We are developing an alternative approach known as Implicit Maximum Likelihood Estimation (IMLE) that maximizes a different lower bound to the marginal log-likelihood without needing to choose a variational distribution and the approximation quality improves with the expressivity of the genrative model. This makes it possible to sidestep the long sampling time of diffusion models, while still maintaining a good approximation to MLE.

Related papers: | | Implicit Maximum Likelihood Estimation | |
Neural Rendering: Popular neural renderers based on 3D Gaussian splatting (3DGS) struggle with post-hoc geometry deformations and motion. Common artifacts include surface tearing or disintegration, and frozen or teleporting parts. This is caused by fundamental limitations of splatting — the shapes of primitives cannot adapt to arbitrary deformations and primitives can hardly move when they are too far from the true position due to vanishing gradients. We are developing an alternative approach known as Proximity Attention Point Rendering (PAPR) that gets around these issues. We reconsider how to form a continuous shape from a discrete point set — rather than filling gaps between points with splats, PAPR interpolates between them using a learned attention kernel. We demonstrated PAPR's capability to learn a point cloud from arbitrary initialization, render at high fidelity under non-rigid post-hoc deformations to the point cloud, and learn large transformations to the point cloud to model scene changes.

Related papers: |
Fast Nearest Neighbour Search (): The method of k-nearest neighbours is widely used in machine learning, statistics, bioinformatics and database systems. Attempts at devising fast algorithms, however, have come up against a recurring obstacle: the curse of dimensionality. Almost all exact algorithms developed over the past 40 years exhibited a time complexity that is exponential in ambient or intrinsic dimensionality, and such persistent failure in overcoming the curse of dimensionality led to conjectures that doing so is impossible. We showed that, surprisingly, this is in fact possible — we developed an exact randomized algorithm whose query time complexity is linear in ambient dimensionality and sublinear in intrinsic dimensionality. The key insight is to avoid the popular strategy of space partitioning, which we argue gives rise to the curse of dimensionality. We demonstrated a speedup of 1-2 orders of magnitude over locality-sensitive hashing (LSH).

Related papers: Fast k-Nearest Neighbour Search via Dynamic Continuous Indexing | Fast k-Nearest Neighbour Search via Prioritized DCI
Learning to Optimize (Slides): While machine learning has been applied to a wide range of domains, one domain that has conspicuously been left untouched is the design of tools that power machine learning itself. In this line of work, we ask the following question: is it possible to automate the design of algorithms used in machine learning? We introduced the first framework for learning a general-purpose iterative optimization algorithm automatically. The key idea is to treat the design of an optimization algorithm as a reinforcement learning/optimal control problem and view a particular update formula (and therefore a particular optimization algorithm) as a particular policy. Finding the optimal policy then corresponds to finding the best optimization algorithm. We parameterize the update formula using a neural net and train it using reinforcement learning to avoid the problem of compounding errors. This has inspired various subsequent work on meta-learning.

Related papers: |

Students

Ph.D. Students

Mehran Aghabozorgi
Mehdi Esmaeilzadeh
Hossein Zaredar

M.Sc. Students

George Shramko
Kian Hosseinkhani

Selected Papers

Generative Modelling

Counterfactual Residual Data Augmentation for Regression
Hossein Mohebbi, Oliver Schulte, Ke Li, Pascal Poupart
International Conference on Machine Learning (ICML), 2026
() ()
Mehran Aghabozorgi, Yanshu Zhang, Alireza Moazeni, Ke Li
International Conference on Learning Representations (ICLR), 2026
() ()
Grayson Lee, Minh Bui, Zhou Shuzi, Yankai Li, Mo Chen, Ke Li
IEEE International Conference on Robotics and Automation (ICRA), 2026
() ()
Yuxiang Fu, Qi Yan, Lele Wang, Ke Li, Renjie Liao
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025
() () ()
Chirag Vashist, Shichong Peng, Ke Li
European Conference on Computer Vision (ECCV), 2024
() () ()
Kiyohiro Nakayama, Mikaela Angelina Uy, Jiahui Huang, Shi-Min Hu, Ke Li, Leonidas J Guibas
IEEE/CVF International Conference on Computer Vision (ICCV), 2023
() () ()
Mehran Aghabozorgi, Shichong Peng, Ke Li
International Conference on Machine Learning (ICML), 2023
() () ()
Shichong Peng, Alireza Moazeni, Ke Li
Advances in Neural Information Processing Systems (NeurIPS), 2022
() ()
Kiarash Zahirnia, Oliver Schulte, Parmis Naddaf, Ke Li
Advances in Neural Information Processing Systems (NeurIPS), 2022
()
Himanshu Arora, Saurabh Mishra, Shichong Peng, Ke Li, Ali Mahdavi-Amiri
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2022
()
Kuan-Chieh Wang, Yan Fu, Ke Li, Ashish Khisti, Richard Zemel, Alireza Makhzani
Advances in Neural Information Processing Systems (NeurIPS), 2021
() ()
Alexia Jolicoeur-Martineau, Ke Li*, R茅mi Pich茅-Taillefer*, Tal Kachman*, Ioannis Mitliagkas
arXiv:2105.14080, 2021
() ()
Shichong Peng, Ke Li
arXiv:2011.01926, 2020
()
Ning Yu, Ke Li, Peng Zhou, Jitendra Malik, Larry Davis, Mario Fritz
European Conference on Computer Vision (ECCV), 2020
Ke Li*, Shichong Peng*, Tianhao Zhang*, Jitendra Malik
International Journal of Computer Vision (IJCV), 2020
(Project Page) () ()
Ke Li*, Tianhao Zhang*, Jitendra Malik
IEEE/CVF International Conference on Computer Vision (ICCV), 2019
(Code) ()
Yedid Hoshen, Ke Li, Jitendra Malik
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019
(Poster)
Ke Li, Jitendra Malik
NeurIPS Workshop on Critiquing and Correcting Trends in Machine Learning, 2018
(Project Page) ()
Ke Li*, Shichong Peng*, Jitendra Malik
arXiv:1810.01406, 2018
(Project Page) (Reviews) (Slides) (Poster) (Code) ()
Ke Li, Jitendra Malik
arXiv:1809.09087, 2018

Neural Rendering

Tackling Misattribution in 3D Intrinsic Decomposition via Proximity Attention Point Rendering
Alireza Moazeni, Shichong Peng, Yanshu Zhang, Chirag Vashist, Ke Li
European Conference on Computer Vision (ECCV), 2026
PointGT: Simultaneous Geometric and Textural Editing for Point-Based Representations
Yanshu Zhang, George Shramko, Pratul P. Srinivasan, Ke Li
European Conference on Computer Vision (ECCV), 2026
P-CORE: Self-Supervised Surface Consistency for Point-Based Neural Editing
Yanshu Zhang, Shichong Peng, Mehran Aghabozorgi, Alireza Moazeni, Ke Li
European Conference on Computer Vision (ECCV), 2026
()
Junru Lin, Chirag Vashist, Mikaela Angelina Uy, Colton Stearns, Xuan Luo, Leonidas Guibas, Ke Li
International Conference on Computer Vision (ICCV), 2025
()
Kiyohiro Nakayama, Mikaela Angelina Uy, Yang You, Ke Li, Leonidas Guibas
Advances in Neural Information Processing Systems (NeurIPS), 2024
() () ()
Shichong Peng, Yanshu Zhang, Ke Li
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (Highlight), 2024
() () ()
Yanshu Zhang*, Shichong Peng*, Alireza Moazeni, Ke Li
Advances in Neural Information Processing Systems (NeurIPS) (Spotlight), 2023
() () ()
Mikaela Angelina Uy, Kiyohiro Nakayama, Guandao Yang, Rahul Krishna Thomas, Leonidas Guibas, Ke Li
Advances in Neural Information Processing Systems (NeurIPS), 2023
() () () () ()
Mikaela Angelina Uy, Ricardo Martin-Brualla, Leonidas Guibas, Ke Li
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023
()
Daniel Rebain, Ke Li, Vincent Sitzmann, Soroosh Yazdani, Kwang Moo Yi, Andrea Tagliasacchi
arXiv:2106.03804, 2021
() ()
Daniel Rebain, Wei Jiang, Soroosh Yazdani, Ke Li, Kwang Moo Yi, Andrea Tagliasacchi
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2021

Learning to Optimize

(Slides) () ()
Ke Li, Jitendra Malik
arXiv:1703.00441, 2017
() (Slides) (Poster) (Code) () ()
Ke Li, Jitendra Malik
arXiv:1606.01885, 2016 and International Conference on Learning Representations (ICLR), 2017

Fast Nearest Neighbour Search

() ()
Yuzhen Mao, Qitong Wang, Martin Ester, Ke Li
International Conference on Learning Representations (ICLR), 2026
() () ()
Yuzhen Mao, Martin Ester, Ke Li
International Conference on Learning Representations (ICLR), 2024
() (Slides) (Project Page) () () (Poster)
Ke Li, Jitendra Malik
International Conference on Machine Learning (ICML), 2017
(Slides) (Project Page) ()
Ke Li, Jitendra Malik
International Conference on Machine Learning (ICML), 2016

Instance Segmentation

Ke Li, Jitendra Malik
European Conference on Computer Vision (ECCV), 2016
Ke Li, Bharath Hariharan, Jitendra Malik
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016

Teaching

CMPT 726: Machine Learning (Spring 2025)

Course Outline

CMPT 983 G200: Generative Models (Fall 2024)

Course Outline

CMPT 983 G200: Generative Models (Spring 2024)

Course Outline

CMPT 726: Machine Learning (Fall 2023)

CMPT 726: Machine Learning (Spring 2023)

CMPT 983 G200: Generative Models (Fall 2022)

CMPT 726: Machine Learning (Spring 2022)

CMPT 983 G200: Generative Models (Fall 2021)

CMPT 726: Machine Learning (Spring 2021)

Lecture Notes