Marc T. Law




NVIDIA Toronto AI Lab
431 King St W, 6th floor, Toronto, ON M5V 3M4 Canada



Introduction I am a senior research scientist at NVIDIA working on machine learning and computer vision. I work in the NVIDIA Toronto AI Lab (Canada) led by Professor Sanja Fidler. I received a PhD in Computer Science from Université Pierre et Marie Curie (now Sorbonne Université in Paris, France) in 2015. My doctoral supervisors were Professor Matthieu Cord and Professor Stéphane Gançarski, I was also supervised by Professor Nicolas Thome. I was a visiting research scholar in the team of Professor Eric Xing at the School of Computer Science, Carnegie Mellon University in 2015~2016. From 2016 to 2019, I was a postdoctoral fellow in the Department of Computer Science (Machine Learning group) at the University of Toronto and Vector Institute under the supervision of Professor Raquel Urtasun and Professor Richard Zemel.
Research Interest My main focus is to propose scalable machine learning methods that can be applied to computer vision tasks.
More specifically, my domains of interest are: distance metric learning, complex representations including hierarchies and graphs, differential geometry, algebraic topology, convex and non-convex optimization, optimization on manifolds, information geometry, structured output prediction, generalization with limited supervision including few-shot learning, semi-supervised learning, active learning, domain adaptation and self-supervised learning.


Motivated students interested in doing an internship in the NVIDIA Toronto AI Lab can send me their resume.

Publications
International Conferences

Self-Supervised Real-to-Sim Scene Generation
Aayush Prakash, Shoubhik Debnath, Jean-Francois Lafleche, Eric Cameracci, Gavriel State, Stan Birchfield, Marc T. Law
IEEE International Conference on Computer Vision (ICCV) 2021
[pdf] [project page]

f-Domain Adversarial Learning: Theory and Algorithms
David Acuna, Guojun Zhang, Marc T. Law, Sanja Fidler
International Conference on Machine Learning (ICML) 2021
[pdf]

Ultrahyperbolic Representation Learning
Marc T. Law, Jos Stam
Neural Information Processing Systems (NeurIPS) 2020
[pdf] [code]

A Theoretical Analysis of the Number of Shots in Few-Shot Learning
Tianshi Cao, Marc T. Law, Sanja Fidler
International Conference on Learning Representations (ICLR) 2020
[pdf]

Video Face Clustering with Unknown Number of Clusters
Makarand Tapaswi, Marc T. Law, Sanja Fidler
IEEE International Conference on Computer Vision (ICCV) 2019
[pdf] [code]

Lorentzian Distance Learning for Hyperbolic Representations
Marc T. Law, Renjie Liao, Jake Snell, Richard S. Zemel
International Conference on Machine Learning (ICML) 2019
[pdf] [code: Retrieval / Binary classification]

Dimensionality Reduction for Representing the Knowledge of Probabilistic Models
Marc T. Law, Jake Snell, Amir-massoud Farahmand, Raquel Urtasun, Richard S. Zemel
International Conference on Learning Representations (ICLR) 2019
[pdf] [code: Visualization / Zero-shot learning 1 2]

Centroid-based Deep Metric Learning for Speaker Recognition
Jixuan Wang, Kuan-Chieh Wang, Marc T. Law, Frank Rudzicz, Michael Brudno
International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2019
[pdf]

Representing Relative Visual Attributes with a Reference-Point-Based Decision Model
Marc T. Law, Paul Weng
International Conference on Pattern Recognition (ICPR) 2018
[pdf]

Deep Spectral Clustering Learning
Marc T. Law, Raquel Urtasun, Richard S. Zemel
International Conference on Machine Learning (ICML) 2017
[pdf] [supp. material] [project page]

Efficient Multiple Instance Metric Learning using Weakly Supervised Data
Marc T. Law, Yiaoliang Yu, Raquel Urtasun, Richard S. Zemel, Eric P. Xing
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2017
[pdf] [code]

Closed-Form Training of Mahalanobis Distance for Supervised Clustering
Marc T. Law, Yiaoliang Yu, Matthieu Cord, Eric P. Xing
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2016
[pdf] [code]

Fantope Regularization in Metric Learning
Marc T. Law, Nicolas Thome, Matthieu Cord
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2014
[doi] [pdf] [code]

Quadruplet-wise Image Similarity Learning
Marc T. Law, Nicolas Thome, Matthieu Cord
IEEE International Conference on Computer Vision (ICCV) 2013
[doi] [pdf]

Structural and Visual Comparisons for Web Page Archiving
Marc T. Law, Nicolas Thome, Stéphane Gançarski, Matthieu Cord
12th edition of the ACM Symposium on Document Engineering (DocEng) 2012
[doi] [pdf] [code] [demo webpage (thanks to Andres Sanoja and Zeynep Pehlivan)]

International Journal

Learning a Distance Metric from Relative Comparisons between Quadruplets of Images
Marc T. Law, Nicolas Thome, Matthieu Cord
International Journal of Computer Vision (IJCV) 2017


International Workshops

Hybrid Pooling Fusion in the BoW Pipeline
Marc T. Law, Nicolas Thome, Matthieu Cord
ECCV 2012 Workshop on Information fusion in Computer Vision for Concept Recognition (ECCV-IFCVCR 2012)
[pdf]

Structural and Visual Similarity Learning for Web Page Archiving
Marc T. Law, Carlos Sureda Gutierrez, Nicolas Thome, Stéphane Gançarski, Matthieu Cord
10th workshop on Content-Based Multimedia Indexing (CBMI) 2012
[pdf]

Book Chapter

Bag-of-Words Image Representation: Key Ideas and Further Insight
Marc T. Law, Nicolas Thome, Matthieu Cord
Fusion in Computer Vision - Understanding Complex Visual Content, Springer 2014

French Conference

Apprentissage de métrique appliqué à la détection de changement de page Web
Marc T. Law, Nicolas Thome, Stéphane Gançarski, Matthieu Cord
CORIA 2015

PhD Thesis

Distance Metric Learning for Image and Webpage Comparison
Université Pierre et Marie Curie (Paris 6), 2015
[pdf] [hal]



Teaching (Moniteur) University Paris 6, Licence 2, LI214: Logic and Automata. (38.5 hours)
University Paris 6, Licence 2, LI215: Imperative programming and data structures in C. (38.5 hours)
University Paris 6, Licence 2, LI230: Object-oriented Programming in Java. (110 hours)
University Paris 6, Master 1 Artificial Intelligence: Project advisor. (3.5 hours)
University Paris 6/Telecom ParisTech, Master 2 Image Processing: Project advisor. (4 hours)