Lingxiao Li

I am a research scientist at Netflix, working on content creation for original series, movies, and video games. Previously, I received my PhD at MIT advised by Justin Solomon in Geometric Data Processing group. Before coming to MIT, I obtained my Bachelor's Degree in Computer Science and Mathematics and my Master's Degree in Mathematics, both at Stanford University.

I am broadly interested in research problems in machine learning, optimization, and computer vision/graphics. I am grateful to have interned at Microsoft Research with Lester Mackey and at Adobe Research with with Noam Aigerman and Vova Kim during my PhD. At Stanford, I was fortunate to have worked with Leonidas Guibas and Daniel Bump, among many others.

Publications

Scalable Methodologies for Optimizing Over Probability Distributions

Lingxiao Li

PhD thesis. MIT, 2024

Thesis

Debiased Distribution Compression

Lingxiao Li, Raaz Dwivedi, Lester Mackey

International Conference on Machine Learning (ICML 2024), Vienna

Paper | Code

Self-Consistent Velocity Matching of Probability Flows

Lingxiao Li, Samuel Hurault, Justin Solomon

Conference on Neural Information Processing Systems (NeurIPS 2023), New Orleans, LA

Paper | Code

Sampling with Mollified Interaction Energy Descent

Lingxiao Li, Qiang Liu, Anna Korba, Mikhail Yurochkin, Justin Solomon

Conference on Learning Representations (ICLR 2023), Kigali

Paper | Code

Learning Proximal Operators to Discover Multiple Optima

Lingxiao Li, Noam Aigerman, Vladimir G. Kim, Jiajin Li, Kristjan Greenewald, Mikhail Yurochkin, Justin Solomon

Conference on Learning Representations (ICLR 2023), Kigali

Paper | Code

Wasserstein Iterative Networks for Barycenter Estimation

Alexander Korotin, Vage Egiazarian, Lingxiao Li, Evgeny Burnaev

Conference on Neural Information Processing Systems (NeurIPS 2022), New Orleans, LA

Paper | Code

Interactive All-Hex Meshing via Cuboid Decomposition

Lingxiao Li, Paul Zhang, Dmitriy Smirnov, Mazdak Abulnaga, Justin Solomon

SIGGRAPH Asia 2021, Tokyo

Paper | Code

Do Neural Optimal Transport Solvers Work? A Continuous Wasserstein-2 Benchmark

Alexander Korotin, Lingxiao Li, Aude Genevay, Justin Solomon, Alexander Filippov, Evgeny Burnaev

Conference on Neural Information Processing Systems (NeurIPS 2021), online

Paper | Code

Large-Scale Wasserstein Gradient Flows

Petr Mokrov*, Alexander Korotin*, Lingxiao Li, Aude Genevay, Justin Solomon, Evgeny Burnaev

Conference on Neural Information Processing Systems (NeurIPS 2021), online

Paper | Code

Continuous Wasserstein-2 Barycenter Estimation without Minimax Optimization

Alexander Korotin, Lingxiao Li, Justin Solomon, Evgeny Burnaev

Conference on Learning Representations (ICLR 2021), online

Paper | Code

Continuous Regularized Wasserstein Barycenters

Lingxiao Li, Aude Genevay, Mikhail Yurochkin, Justin Solomon

Conference on Neural Information Processing Systems (NeurIPS 2020), online

Paper | Code

Supervised Fitting of Geometric Primitives to 3D Point Clouds

Lingxiao Li*, Minhyuk Sung*, Anastasia Dubrovina, Li Yi, Leonidas Guibas

Oral Presentation
Conference on Computer Vision and Pattern Recognition (CVPR 2019), Long Beach, CA

Paper | Code | Video

Branching Rules of Classical Lie Groups in Two Ways

Lingxiao Li

Undergraduate honors thesis. Stanford University, 2018

Thesis