Yanshuai Cao
Research Team Lead
PhD Computer Science, University of Toronto
Yanshuai Cao is currently a research team lead at Borealis AI. His long-term research goal is to create machines that can learn from as little supervision as possible and as quickly as humans do when facing new data. Currently, his research spans generative models, continual learning, computer vision and natural language processing. Previously, he also explored Bayesian nonparametric methods to that same end.
Yanshuai received his PhD in Computer Science from the Department of Computer Science at the University of Toronto, where he was advised by Professor David J. Fleet and Professor Aaron Hertzmann.
Research Areas
Computer Vision
Natural Language Processing
Publications
July 12, 2020
Evaluating Lossy Compression Rates of Deep Generative Models
International Conference on Machine Learning (ICML), 2020
July 12, 2020
On Variational Learning of Controllable Representations for Text without Supervision
International Conference on Machine Learning (ICML), 2020
June 3, 2020
Better Long-Range Dependency by Bootstrapping A Mutual Information Regularizer
International Conference on Artificial Intelligence and Statistics (AISTATS), 2020
July 28, 2019
A Cross-Domain Transferable Neural Coherence Model
Association for Computational Linguistics (ACL), 2019
Dec. 3, 2018
Few-Shot Self Reminder to Overcome Catastrophic Forgetting
Workshop on Continual Learning (NeurIPS), 2018
Dec. 3, 2018
Compositional Hard Negatives for Visual Semantic Embeddings via an Adversary
Workshop on ViGIL (NeurIPS), 2018
July 15, 2018
Adversarial Contrastive Estimation
Association for Computational Linguistics (ACL), 2018
April 30, 2018
Improving GAN Training via Binarized Representation Entropy (BRE) Regularization
International Conference on Learning Representations (ICLR), 2018
Aug. 6, 2017
Implicit Manifold Learning on Generative Adversarial Networks
Workshop on Implicit Models (ICML), 2017