no code implementations • 25 Apr 2024 • Nathanaël Perraudin, Adrien Teutrie, Cécile Hébert, Guillaume Obozinski
We consider the problem of regularized Poisson Non-negative Matrix Factorization (NMF) problem, encompassing various regularization terms such as Lipschitz and relatively smooth functions, alongside linear constraints.
1 code implementation • 14 Dec 2023 • Andreas Bergmeister, Karolis Martinkus, Nathanaël Perraudin, Roger Wattenhofer
However, most existing methods struggle with large graphs due to the complexity of representing the entire joint distribution across all node pairs and capturing both global and local graph structures simultaneously.
no code implementations • 20 Feb 2023 • Stefania Russo, Nathanaël Perraudin, Steven Stalder, Fernando Perez-Cruz, Joao Paulo Leitao, Guillaume Obozinski, Jan Dirk Wegner
In this technical report we compare different deep learning models for prediction of water depth rasters at high spatial resolution.
1 code implementation • 4 Oct 2022 • Kilian Konstantin Haefeli, Karolis Martinkus, Nathanaël Perraudin, Roger Wattenhofer
Denoising diffusion probabilistic models and score-matching models have proven to be very powerful for generative tasks.
1 code implementation • 23 May 2022 • Steven Stalder, Nathanaël Perraudin, Radhakrishna Achanta, Fernando Perez-Cruz, Michele Volpi
These attributions are provided in the form of masks that only show the classifier-relevant parts of an image, masking out the rest.
1 code implementation • 4 Apr 2022 • Karolis Martinkus, Andreas Loukas, Nathanaël Perraudin, Roger Wattenhofer
We approach the graph generation problem from a spectral perspective by first generating the dominant parts of the graph Laplacian spectrum and then building a graph matching these eigenvalues and eigenvectors.
no code implementations • 24 Sep 2021 • Romana Rust, Achilleas Xydis, Kurt Heutschi, Nathanaël Perraudin, Gonzalo Casas, Chaoyu Du, Jürgen Strauss, Kurt Eggenschwiler, Fernando Perez-Cruz, Fabio Gramazio, Matthias Kohler
In this paper, we present the automated data acquisition setup, the data processing and the computational generation of diffusive surface structures.
9 code implementations • ICLR 2020 • Michaël Defferrard, Martino Milani, Frédérick Gusset, Nathanaël Perraudin
DeepSphere, a method based on a graph representation of the sampled sphere, strikes a controllable balance between these two desiderata.
1 code implementation • 14 Oct 2020 • Karolis Martinkus, Aurelien Lucchi, Nathanaël Perraudin
However, the dynamics of many real-world systems are challenging to learn due to the presence of nonlinear potentials and a number of interactions that scales quadratically with the number of particles $N$, as in the case of the N-body problem.
2 code implementations • 11 May 2020 • Andres Marafioti, Piotr Majdak, Nicki Holighaus, Nathanaël Perraudin
We introduce GACELA, a generative adversarial network (GAN) designed to restore missing musical audio data with a duration ranging between hundreds of milliseconds to a few seconds, i. e., to perform long-gap audio inpainting.
no code implementations • 17 Apr 2020 • Nathanaël Perraudin, Sandro Marcon, Aurelien Lucchi, Tomasz Kacprzak
Weak gravitational lensing mass maps play a crucial role in understanding the evolution of structures in the universe and our ability to constrain cosmological models.
1 code implementation • 15 Aug 2019 • Nathanaël Perraudin, Ankit Srivastava, Aurelien Lucchi, Tomasz Kacprzak, Thomas Hofmann, Alexandre Réfrégier
Our results show that the proposed model produces samples of high visual quality, although the statistical analysis reveals that capturing rare features in the data poses significant problems for the generative models.
no code implementations • 31 May 2019 • Younjoo Seo, Andreas Loukas, Nathanaël Perraudin
This paper focuses on the discrimination capacity of aggregation functions: these are the permutation invariant functions used by graph neural networks to combine the features of nodes.
6 code implementations • 8 Apr 2019 • Michaël Defferrard, Nathanaël Perraudin, Tomasz Kacprzak, Raphael Sgier
Spherical data is found in many applications.
1 code implementation • 36th International Conference on Machine Learning 2019 • Andrés Marafioti, Nicki Holighaus, Nathanaël Perraudin, Piotr Majdak
We demonstrate the potential of deliberate generative TF modeling by training a generative adversarial network (GAN) on short-time Fourier features.
5 code implementations • 29 Oct 2018 • Nathanaël Perraudin, Michaël Defferrard, Tomasz Kacprzak, Raphael Sgier
We present a spherical CNN for analysis of full and partial HEALPix maps, which we call DeepSphere.
1 code implementation • 29 Oct 2018 • Andrés Marafioti, Nicki Holighaus, Piotr Majdak, Nathanaël Perraudin
We studied the ability of deep neural networks (DNNs) to restore missing audio content based on its context, a process usually referred to as audio inpainting.
no code implementations • ICLR 2019 • Vassilis Kalofolias, Nathanaël Perraudin
In this paper, we show how to scale it, obtaining an approximation with leading cost of $\mathcal{O}(n\log(n))$, with quality that approaches the exact graph learning model.
no code implementations • 5 May 2017 • Francesco Grassi, Andreas Loukas, Nathanaël Perraudin, Benjamin Ricaud
An emerging way to deal with high-dimensional non-euclidean data is to assume that the underlying structure can be captured by a graph.
no code implementations • 19 Feb 2017 • Johan Paratte, Nathanaël Perraudin, Pierre Vandergheynst
Visualizing high-dimensional data has been a focus in data analysis communities for decades, which has led to the design of many algorithms, some of which are now considered references (such as t-SNE for example).
no code implementations • 1 Nov 2016 • Andreas Loukas, Nathanaël Perraudin
This paper considers regression tasks involving high-dimensional multivariate processes whose structure is dependent on some {known} graph topology.
no code implementations • 11 Jan 2016 • Nathanaël Perraudin, Pierre Vandergheynst
Graphs are a central tool in machine learning and information processing as they allow to conveniently capture the structure of complex datasets.