CAII Spring Seminar: "Tensor Methods for Deep Learning with TensorLy and PyTorch"

Event Type

Seminar/Symposium

Sponsor

Center for Artificial Intelligence Innovation

Date

Apr 5, 2021   11:00 am - 12:00 pm  

Speaker

Dr. Jean Kossaifi, Senior Research Scientist at NVIDIA

Views

146

Originating Calendar

Center for Artificial Intelligence Innovation

Jean Kossaifi , Senior Research Scientist at NVIDIA will give a presentation during the CAII Seminar Series on Monday, April 5 at 11:00 a.m. The talk is titled “Tensor Methods for Deep Learning with TensorLy and PyTorch". 

View Seminar here: https://go.ncsa.illinois.edu/CAIISpringSemesterSeriesSP21  

Abstract:
Tensor methods are a natural extension of matrix algebraic methods to higher orders. While they have been extensively used in fields such as quantum systems, signal processing and chemometrics, they only recently started to be employed in machine learning. They have the potential to both help us develop better theory about deep learning, and to leverage the multi-dimensional structure in the data. This ability to leverage topological structure can be crucial when learning from spatio-temporal data or from structured data such as MRI.

In this presentation, I will give an overview of tensor methods for deep learning, starting with a short introduction to tensor decomposition, how to leverage tensor methods to design better deep models, for improved performance or speed, model compression and robustness and finally practical implementation using TensorLy and TensorLy-Torch.

Speaker Bio: 
Dr. Jean Kossaifi is a Senior Research Scientist at NVIDIA. His current focus is on tensor methods for machine learning. Particularly, efficient combination of these methods with deep learning to develop better models that are memory and computation efficient, while being more robust to noise, random or adversarial, as well as domain shift. He is the creator of TensorLy, a high-level API for tensor methods and deep tensorized neural networks in Python, designed to make tensor learning simple and accessible. He has also worked extensively on face analysis and facial affect estimation in naturalistic conditions, a field which bridges the gap between computer vision and machine learning. 

Prior to joining NVIDIA, Jean worked at the Samsung AI Center in Cambridge. He received his PhD and MSc from Imperial College London, where he worked with Prof. Maja Pantic. He also holds a French Engineering Diploma / MSc in Applied Mathematics, Computing and Finance and obtained a BSc in advanced mathematics in parallel.