1 code implementation • 22 Feb 2023 • Song Duong, Alberto Lumbreras, Mike Gartrell, Patrick Gallinari
Our model is designed to handle the tasks of D2T and T2D jointly.
no code implementations • 18 Jan 2023 • Carlo Metta, Riccardo Guidotti, Yuan Yin, Patrick Gallinari, Salvatore Rinzivillo
We propose a use case study, for skin lesion diagnosis, illustrating how it is possible to provide the practitioner with explanations on the decisions of a state of the art deep neural network classifier trained to characterize skin lesions from examples.
1 code implementation • 15 Dec 2022 • Florent Bonnet, Ahmed Jocelyn Mazari, Paola Cinnella, Patrick Gallinari
Surrogate models are necessary to optimize meaningful quantities in physical dynamics as their recursive numerical resolutions are often prohibitively expensive.
no code implementations • 15 Nov 2022 • Jingang Qu, Thibault Faney, Ze Wang, Patrick Gallinari, Soleiman Yousef, Jean-Charles de Hemptinne
HMOE uses hypernetworks taking vectors as input to generate experts' weights, which allows experts to share useful meta-knowledge and enables exploring experts' similarities in a low-dimensional vector space.
no code implementations • 3 Oct 2022 • Skander Karkar, Ibrahim Ayed, Emmanuel de Bézenac, Patrick Gallinari
Greedy layer-wise or module-wise training of neural networks is compelling in constrained and on-device settings, as it circumvents a number of problems of end-to-end back-propagation.
1 code implementation • 29 Sep 2022 • Yuan Yin, Matthieu Kirchmeyer, Jean-Yves Franceschi, Alain Rakotomamonjy, Patrick Gallinari
Effective data-driven PDE forecasting methods often rely on fixed spatial and / or temporal discretizations.
1 code implementation • 29 Jun 2022 • Léon Migus, Yuan Yin, Jocelyn Ahmed Mazari, Patrick Gallinari
In this work, we study three multi-resolution schema with integral kernel operators that can be approximated with \emph{Message Passing Graph Neural Networks} (MPGNNs).
1 code implementation • 29 Jun 2022 • Florent Bonnet, Jocelyn Ahmed Mazari, Thibaut Munzer, Pierre Yser, Patrick Gallinari
Recent progress in \emph{Geometric Deep Learning} (GDL) has shown its potential to provide powerful data-driven models.
2 code implementations • 19 May 2022 • Alexandre Ramé, Matthieu Kirchmeyer, Thibaud Rahier, Alain Rakotomamonjy, Patrick Gallinari, Matthieu Cord
Standard neural networks struggle to generalize under distribution shifts in computer vision.
no code implementations • 6 May 2022 • Jingang Qu, Thibault Faney, Jean-Charles de Hemptinne, Soleiman Yousef, Patrick Gallinari
Phase equilibrium calculations are an essential part of numerical simulations of multi-component multi-phase flow in porous media, accounting for the largest share of the computational time.
1 code implementation • 1 Feb 2022 • Matthieu Kirchmeyer, Yuan Yin, Jérémie Donà, Nicolas Baskiotis, Alain Rakotomamonjy, Patrick Gallinari
Data-driven approaches to modeling physical systems fail to generalize to unseen systems that share the same general dynamics with the learning domain, but correspond to different physical contexts.
no code implementations • 22 Nov 2021 • Carlo Metta, Andrea Beretta, Riccardo Guidotti, Yuan Yin, Patrick Gallinari, Salvatore Rinzivillo, Fosca Giannotti
A key issue in critical contexts such as medical diagnosis is the interpretability of the deep learning models adopted in decision-making systems.
1 code implementation • ICLR 2022 • Matthieu Kirchmeyer, Alain Rakotomamonjy, Emmanuel de Bezenac, Patrick Gallinari
We consider the problem of unsupervised domain adaptation (UDA) between a source and a target domain under conditional and label shift a. k. a Generalized Target Shift (GeTarS).
no code implementations • ICLR 2022 • Jérémie Dona, Marie Déchelle, Patrick Gallinari, Marina Levy
A common practice to identify the respective parameters of the physical and ML components is to formulate the problem as supervised learning on observed trajectories.
no code implementations • EMNLP 2021 • Bruno Taillé, Vincent Guigue, Geoffrey Scoutheeten, Patrick Gallinari
State-of-the-art NLP models can adopt shallow heuristics that limit their generalization capability (McCoy et al., 2019).
1 code implementation • 16 Sep 2021 • Matthieu Kirchmeyer, Patrick Gallinari, Alain Rakotomamonjy, Amin Mantrach
Moreover, we compare the target error of our Adaptation-imputation framework and the "ideal" target error of a UDA classifier without missing target components.
no code implementations • 21 Jul 2021 • Jérémie Dona, Patrick Gallinari
The second one corresponds to the elimination of redundant information by selecting variables in a correlated fashion which requires modeling the covariance of the binary distribution.
1 code implementation • 10 Jun 2021 • Jean-Yves Franceschi, Emmanuel de Bézenac, Ibrahim Ayed, Mickaël Chen, Sylvain Lamprier, Patrick Gallinari
We propose a novel theoretical framework of analysis for Generative Adversarial Networks (GANs).
1 code implementation • NeurIPS 2021 • Yuan Yin, Ibrahim Ayed, Emmanuel de Bézenac, Nicolas Baskiotis, Patrick Gallinari
Both are sub-optimal: the former disregards the discrepancies between environments leading to biased solutions, while the latter does not exploit their potential commonalities and is prone to scarcity problems.
2 code implementations • EMNLP 2021 • Clément Rebuffel, Thomas Scialom, Laure Soulier, Benjamin Piwowarski, Sylvain Lamprier, Jacopo Staiano, Geoffrey Scoutheeten, Patrick Gallinari
QuestEval is a reference-less metric used in text-to-text tasks, that compares the generated summaries directly to the source text, by automatically asking and answering questions.
2 code implementations • EMNLP 2021 • Thomas Scialom, Paul-Alexis Dray, Patrick Gallinari, Sylvain Lamprier, Benjamin Piwowarski, Jacopo Staiano, Alex Wang
Summarization evaluation remains an open research problem: current metrics such as ROUGE are known to be limited and to correlate poorly with human judgments.
1 code implementation • 4 Feb 2021 • Clément Rebuffel, Marco Roberti, Laure Soulier, Geoffrey Scoutheeten, Rossella Cancelliere, Patrick Gallinari
Specifically, we propose a Multi-Branch Decoder which is able to leverage word-level labels to learn the relevant parts of each training instance.
Ranked #3 on
Table-to-Text Generation
on WikiBio
no code implementations • NeurIPS 2020 • Emmanuel de Bézenac, Syama Sundar Rangapuram, Konstantinos Benidis, Michael Bohlke-Schneider, Richard Kurle, Lorenzo Stella, Hilaf Hasson, Patrick Gallinari, Tim Januschowski
This paper tackles the modelling of large, complex and multivariate time series panels in a probabilistic setting.
3 code implementations • 24 Nov 2020 • Manon Césaire, Lucas Schott, Hatem Hajri, Sylvain Lamprier, Patrick Gallinari
This paper introduces stochastic sparse adversarial attacks (SSAA), standing as simple, fast and purely noise-based targeted and untargeted attacks of neural network classifiers (NNC).
no code implementations • 13 Nov 2020 • Patrick Bordes, Eloi Zablocki, Benjamin Piwowarski, Patrick Gallinari
We show the efficiency of our Cross-Modal CycleGAN model (CM-GAN) on the ImageNet T-ZSL task where we obtain state-of-the-art results.
1 code implementation • INLG (ACL) 2020 • Clément Rebuffel, Laure Soulier, Geoffrey Scoutheeten, Patrick Gallinari
Evaluations on the widely used WikiBIO and WebNLG benchmarks demonstrate the effectiveness of this framework compared to state-of-the-art models.
2 code implementations • ICLR 2021 • Yuan Yin, Vincent Le Guen, Jérémie Dona, Emmanuel de Bézenac, Ibrahim Ayed, Nicolas Thome, Patrick Gallinari
In this work, we introduce the APHYNITY framework, a principled approach for augmenting incomplete physical dynamics described by differential equations with deep data-driven models.
3 code implementations • EMNLP 2020 • Bruno Taillé, Vincent Guigue, Geoffrey Scoutheeten, Patrick Gallinari
Despite efforts to distinguish three different evaluation setups (Bekoulis et al., 2018), numerous end-to-end Relation Extraction (RE) articles present unreliable performance comparison to previous work.
1 code implementation • 17 Sep 2020 • Skander Karkar, Ibrahim Ayed, Emmanuel de Bézenac, Patrick Gallinari
From this observation, we reformulate the learning problem as follows: finding neural networks which solve the task while transporting the data as efficiently as possible.
1 code implementation • ICLR 2021 • Jérémie Donà, Jean-Yves Franceschi, Sylvain Lamprier, Patrick Gallinari
A recent line of work in the machine learning community addresses the problem of predicting high-dimensional spatiotemporal phenomena by leveraging specific tools from the differential equations theory.
no code implementations • INLG (ACL) 2020 • Thomas Scialom, Patrick Bordes, Paul-Alexis Dray, Jacopo Staiano, Patrick Gallinari
Pre-trained language models have recently contributed to significant advances in NLP tasks.
1 code implementation • ICML 2020 • Jean-Yves Franceschi, Edouard Delasalles, Mickaël Chen, Sylvain Lamprier, Patrick Gallinari
Designing video prediction models that account for the inherent uncertainty of the future is challenging.
Ranked #1 on
Video Prediction
on KTH 64x64 cond10 pred30
no code implementations • IJCNLP 2019 • Patrick Bordes, Eloi Zablocki, Laure Soulier, Benjamin Piwowarski, Patrick Gallinari
To overcome this limitation, we propose to transfer visual information to textual representations by learning an intermediate representation space: the grounded space.
1 code implementation • 22 Jan 2020 • Bruno Taillé, Vincent Guigue, Patrick Gallinari
Contextualized embeddings use unsupervised language model pretraining to compute word representations depending on their context.
1 code implementation • 20 Dec 2019 • Clément Rebuffel, Laure Soulier, Geoffrey Scoutheeten, Patrick Gallinari
This however loses most of the structure contained in the data.
1 code implementation • 18 Dec 2019 • Arthur Pajot, Emmanuel de Bezenac, Patrick Gallinari
This allows us sampling from the latent component in order to generate a distribution of images associated to an observation.
no code implementations • 25 Sep 2019 • Yuan Yin, Arthur Pajot, Emmanuel de Bézenac, Patrick Gallinari
We tackle the problem of inpainting occluded area in spatiotemporal sequences, such as cloud occluded satellite observations, in an unsupervised manner.
no code implementations • 25 Sep 2019 • Matthieu Kirchmeyer, Patrick Gallinari, Alain Rakotomamonjy, Amin Mantrach
Motivated by practical applications, we consider unsupervised domain adaptation for classification problems, in the presence of missing data in the target domain.
no code implementations • 4 Jun 2019 • Emmanuel de Bézenac, Ibrahim Ayed, Patrick Gallinari
Domain Translation is the problem of finding a meaningful correspondence between two domains.
1 code implementation • 30 May 2019 • Karan Aggarwal, Matthieu Kirchmeyer, Pranjul Yadav, S. Sathiya Keerthi, Patrick Gallinari
Such a real world situation is best represented using an implicit model in which an extra noise vector, $z$ is included with $x$ as input.
1 code implementation • ICLR 2019 • Arthur Pajot, Emmanuel de Bezenac, Patrick Gallinari
We address the problem of recovering an underlying signal from lossy, inaccurate observations in an unsupervised setting.
no code implementations • ICLR 2019 • Ibrahim Ayed, Emmanuel de Bézenac, Arthur Pajot, Patrick Gallinari
Spatio-Temporal processes bear a central importance in many applied scientific fields.
1 code implementation • 26 Apr 2019 • Marco Roberti, Giovanni Bonetta, Rossella Cancelliere, Patrick Gallinari
In the last few years, many different methods have been focusing on using deep recurrent neural networks for natural language generation.
Ranked #2 on
Data-to-Text Generation
on E2E NLG Challenge
no code implementations • 24 Apr 2019 • Eloi Zablocki, Patrick Bordes, Benjamin Piwowarski, Laure Soulier, Patrick Gallinari
Zero-Shot Learning (ZSL) aims at classifying unlabeled objects by leveraging auxiliary knowledge, such as semantic representations.
no code implementations • 26 Feb 2019 • Ibrahim Ayed, Emmanuel de Bézenac, Arthur Pajot, Julien Brajard, Patrick Gallinari
We consider the problem of forecasting complex, nonlinear space-time processes when observations provide only partial information of on the system's state.
no code implementations • JEPTALNRECITAL 2018 • Charles-Emmanuel Dias, Clara de Forsan de Gainon Gabriac, Patrick Gallinari, Vincent Guigue
Dans le cadre de l{'}atelier DEFT 2018 nous nous sommes int{\'e}ress{\'e}s {\`a} la classification de microblogs (ici, des tweets) r{\'e}dig{\'e}s en fran{\c{c}}ais.
no code implementations • 23 Apr 2018 • Ali Ziat, Edouard Delasalles, Ludovic Denoyer, Patrick Gallinari
We introduce a dynamical spatio-temporal model formalized as a recurrent neural network for forecasting time series of spatial processes, i. e. series of observations sharing temporal and spatial dependencies.
1 code implementation • 20 Dec 2017 • Wenjie Zheng, Aurélien Bellet, Patrick Gallinari
We consider the problem of learning a high-dimensional but low-rank matrix from a large-scale dataset distributed over several machines, where low-rankness is enforced by a convex trace norm constraint.
2 code implementations • ICLR 2018 • Emmanuel de Bezenac, Arthur Pajot, Patrick Gallinari
We consider the use of Deep Learning methods for modeling complex phenomena like those occurring in natural physical processes.
no code implementations • 9 Nov 2017 • Éloi Zablocki, Benjamin Piwowarski, Laure Soulier, Patrick Gallinari
Representing the semantics of words is a long-standing problem for the natural language processing community.
no code implementations • 25 Aug 2015 • Rossella Cancelliere, Mario Gai, Patrick Gallinari, Luca Rubini
In this paper we consider the training of single hidden layer neural networks by pseudoinversion, which, in spite of its popularity, is sometimes affected by numerical instability issues.
no code implementations • 17 Dec 2014 • Mickaël Poussevin, Vincent Guigue, Patrick Gallinari
We propose to augment rating based recommender systems by providing the user with additional information which might help him in his choice or in the understanding of the recommendation.
no code implementations • 2 Oct 2014 • Ludovic Denoyer, Patrick Gallinari
Instead of considering global transformations, like in classical multilayer networks, this model allows us for learning a set of local transformations.
no code implementations • 20 Dec 2013 • Cédric Lagnier, Simon Bourigault, Sylvain Lamprier, Ludovic Denoyer, Patrick Gallinari
We introduce a model for predicting the diffusion of content information on social media.
no code implementations • 20 Dec 2013 • Gabriella Contardo, Ludovic Denoyer, Thierry Artieres, Patrick Gallinari
We propose to deal with sequential processes where only partial observations are available by learning a latent representation space on which policies may be accurately learned.
no code implementations • 20 Dec 2013 • Gabriel Dulac-Arnold, Ludovic Denoyer, Nicolas Thome, Matthieu Cord, Patrick Gallinari
In this paper, we investigate a new framework for image classification that adaptively generates spatial representations.
no code implementations • NeurIPS 2013 • Moustapha M. Cisse, Nicolas Usunier, Thierry Artières, Patrick Gallinari
This paper presents an approach to multilabel classification (MLC) with a large number of labels.