Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Picture for François Fleuret

François Fleuret

Idiap Research Institute, EPFL

Language Models are Few-Shot Butlers


Apr 16, 2021
Vincent Micheli, François Fleuret


  Access Paper or Ask Questions

Structural analysis of an all-purpose question answering model


Apr 13, 2021
Vincent Micheli, Quentin Heinrich, François Fleuret, Wacim Belblidia


  Access Paper or Ask Questions

Unsupervised clustering of series using dynamic programming and neural processes


Jan 26, 2021
Karthigan Sinnathamby, Chang-Yu Hou, Lalitha Venkataramanan, Vasileios-Marios Gkortsas, François Fleuret


  Access Paper or Ask Questions

Unsupervised clustering of series using dynamic programming


Jan 23, 2021
Karthigan Sinnathamby, Chang-Yu Hou, Lalitha Venkataramanan, Vasileios-Marios Gkortsas, François Fleuret


  Access Paper or Ask Questions

On the importance of pre-training data volume for compact language models


Oct 09, 2020
Vincent Micheli, Martin d'Hoffschmidt, François Fleuret

* EMNLP 2020; typo corrected 

  Access Paper or Ask Questions

Fast Transformers with Clustered Attention


Jul 09, 2020
Apoorv Vyas, Angelos Katharopoulos, François Fleuret


  Access Paper or Ask Questions

Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention


Jun 30, 2020
Angelos Katharopoulos, Apoorv Vyas, Nikolaos Pappas, François Fleuret

* ICML 2020, project at https://linear-transformers.com/ 

  Access Paper or Ask Questions

Multi-task Reinforcement Learning with a Planning Quasi-Metric


Feb 08, 2020
Vincent Micheli, Karthigan Sinnathamby, François Fleuret


  Access Paper or Ask Questions

On the Tunability of Optimizers in Deep Learning


Oct 25, 2019
Prabhu Teja Sivaprasad, Florian Mai, Thijs Vogels, Martin Jaggi, François Fleuret

* Under review at ICLR 2020. 16 pages, 8 figures 

  Access Paper or Ask Questions

Processing Megapixel Images with Deep Attention-Sampling Models


May 03, 2019
Angelos Katharopoulos, François Fleuret

* To appear in ICML 2019 

  Access Paper or Ask Questions

Reducing Noise in GAN Training with Variance Reduced Extragradient


Apr 18, 2019
Tatjana Chavdarova, Gauthier Gidel, François Fleuret, Simon Lacoste-Julien


  Access Paper or Ask Questions

Not All Samples Are Created Equal: Deep Learning with Importance Sampling


Jun 09, 2018
Angelos Katharopoulos, François Fleuret

* Accepted at ICML 2018 (short oral) 

  Access Paper or Ask Questions

SGAN: An Alternative Training of Generative Adversarial Networks


Dec 06, 2017
Tatjana Chavdarova, François Fleuret


  Access Paper or Ask Questions

Biased Importance Sampling for Deep Neural Network Training


Sep 13, 2017
Angelos Katharopoulos, François Fleuret


  Access Paper or Ask Questions

The WILDTRACK Multi-Camera Person Dataset


Jul 28, 2017
Tatjana Chavdarova, Pierre Baqué, Stéphane Bouquet, Andrii Maksai, Cijo Jose, Louis Lettry, Pascal Fua, Luc Van Gool, François Fleuret


  Access Paper or Ask Questions

Deep Multi-camera People Detection


Jul 23, 2017
Tatjana Chavdarova, François Fleuret


  Access Paper or Ask Questions

Deep Occlusion Reasoning for Multi-Camera Multi-Target Detection


Apr 20, 2017
Pierre Baqué, François Fleuret, Pascal Fua


  Access Paper or Ask Questions

A Sub-Quadratic Exact Medoid Algorithm


Apr 12, 2017
James Newling, François Fleuret

* Version 2: Added acknowledgements, Version 3: Post-acceptance at AISTATS 2017, Version 4: N-1 -> N denominator correction 

  Access Paper or Ask Questions

Social Scene Understanding: End-to-End Multi-Person Action Localization and Collective Activity Recognition


Nov 28, 2016
Timur Bagautdinov, Alexandre Alahi, François Fleuret, Pascal Fua, Silvio Savarese


  Access Paper or Ask Questions

Predicting the dynamics of 2d objects with a deep residual network


Nov 24, 2016
François Fleuret


  Access Paper or Ask Questions

Multi-Modal Mean-Fields via Cardinality-Based Clamping


Nov 23, 2016
Pierre Baqué, François Fleuret, Pascal Fua

* Submitted for review to CVPR 2017 

  Access Paper or Ask Questions

Nested Mini-Batch K-Means


Sep 12, 2016
James Newling, François Fleuret

* Nested Mini-Batch K-Means, Proceedings of the International Conference on Neural Information Processing Systems (NIPS), 2016 
* 8 pages + Supplementary Material. Version 2 : new experiments added. Version 3 : Add acknowledgments, upper case in title. Version 4 : Correct spelling of Acknowledgements, change title. Version 5: camera ready NIPS 

  Access Paper or Ask Questions

Fast K-Means with Accurate Bounds


Sep 11, 2016
James Newling, François Fleuret

* Proceedings of the International Conference on Machine Learning (ICML) pp. 936-944, 2016 
* 8 pages + supplementary material v2: mlpack installed with optimisation (previously installed in DEBUG) v3: Annulus -> Annular v4: Author affiliation update v5: Synced with version at ICML, now including Suppl. Mat 

  Access Paper or Ask Questions

Principled Parallel Mean-Field Inference for Discrete Random Fields


Dec 03, 2015
Pierre Baqué, Timur Bagautdinov, François Fleuret, Pascal Fua

* The first two authors contributed equally 

  Access Paper or Ask Questions

A provably convergent alternating minimization method for mean field inference


Feb 20, 2015
Pierre Baqué, Jean-Hubert Hours, François Fleuret, Pascal Fua

* Submitted to Colt 2015 

  Access Paper or Ask Questions

Occam's hammer: a link between randomized learning and multiple testing FDR control


Aug 29, 2006
Gilles Blanchard, François Fleuret

* 13 pages -- conference communication type format 

  Access Paper or Ask Questions