Jean Kossaifi

40 updates

Talk · Keynote

AI+Science: Accelerating Discovery, Stanford University

Talk · Invited Talk

Neural Operators: Applications

HACE 2026 Workshop on HPC/AI Hybridization, IRIT, Toulouse

Invited applications talk on neural operators in a session examining AI-based surrogate models for solving partial differential equations.

News

NeuralOperator joins the PyTorch Ecosystem: Learning in Infinite Dimension with Neural Operators

Talk · Invited

Tensor Decompositions in Modern AI

TRICAP 2025, Ålesund, Norway

Invited talk at this interdisciplinary meeting on tensor decompositions and algorithms, discussing their increasing relevance in modern AI.

News

Co-organized workshop with Topal team at INRIA Bordeaux on efficient scaling of neural architectures, covering re-materialization, offloading, scheduling and model pipelining

News

Co-organizer, ICML Workshop on Advancing Neural Network Training (WANT): Computational Efficiency, Scalability, and Resource Optimization

News

Co-authored a book chapter, "Tensor methods in deep learning", in the book Signal Processing and Machine Learning Theory

News

Invited speaker at the Scale by the Bay, bay area AI, 2023, hosted by IBM on AI for Science with Neural Operators

News

New blog post on Weather Modeling with Spherical Neural Operators

News

Organizer, NeurIPS Workshop on Advancing Neural Network Training (WANT): Computational Efficiency, Scalability, and Resource Optimization

News

We set a new world record, for the largest quantum circuit simulation, using TensorLy-Quantum, NVIDIA's cuQuantum library and a new methodology we developed. Using 896 GPUs to simulate 1,688 qubits, we were able to solve the MaxCut problem for a graph with 3,375 vertices!

News

New NVIDIA blog post on Tensor Methods for Deep Learning

News Highlighted update.

Our paper on emotion analysis in the wild was published in Nature Machine Intelligence: Estimation of continuous valence and arousal levels from faces in naturalistic conditions. Our method jointly estimates categorical emotions and continuous valence and arousal. It was the first system to match or outperform expert human annotators on continuous valence and arousal estimation in the wild: on AffectNet and SEWA, its agreement with the reference annotations matched or exceeded the agreement between expert annotators.

News

Co-organizer, NeurIPS Second Workshop on Quantum Tensor Networks in Machine Learning

News

Accepted paper at CVPR 2020: Factorized Higher-Order CNNs with an Application to Spatio-Temporal Emotion Estimation

News

New AAAI paper: Incremental multi-domain learning with network latent tensor factorization

News

New book on Spectral Learning on Matrices and Tensors published

News

Invited speaker at the Third International Workshop on “Robust Subspace Learning and Applications in Computer Vision” at ICCV 2019

News Highlighted update.

Paper SEWA DB: A Rich Database for Audio-Visual Emotion and Sentiment Research in the Wild accepted at TPAMI!

News

Delivered a tutorial on advanced deep learning, tensor methods and quantization at Caltech

News

Delivered a GTC talk “Take Your Machine Learning to Higher Dimensions with Tensor Methods”, together with my colleague Chris Choy

News

Paper on T-Net: Parametrizing Fully Convolutional Nets with a Single High-Order Tensor accepted at CVPR'19

News

Delivered a tutorial on Deep Learning and Tensor Methods at Caltech

News

Taught the class on Tensor Methods for Large Scale Machine-Learning at the IfI Summer School 2018 on Machine Learning

News

New paper accepted at IEEE CVPR on Geometry-Aware Generative Adversarial Networks

News Highlighted update.

Best paper award at the NIPS MLtrain workshop for Tensor Contraction & Regression Networks with TensorLy

News

Co-organizing the ICCV 2017 Workshop on Matrix and Tensor Factorization Methods for Computer Vision

News

Paper on Tensor Contraction Layers for Parsimonious Deep Nets accepted at CVPR'17 Workshop on Tensor Methods in Computer Vision