Nicolas Courty
@nicolas_courty
Followers
569
Following
1K
Media
24
Statuses
218
Head of Obelix group @irisa Machine Learning, Optimal Transport and Remote sensing
Joined October 2017
RT @SibylleMarcotte: 🏆Didn't get the Physics Nobel Prize this year, but really excited to share that I've been named one of the #FWIS2024 @….
0
20
0
RT @Bjoern_Michele: 🚨 Interested in domain adaptation and generalization for 3D data? 🚨. It’s tough to keep up with all the new amazing wor….
0
7
0
RT @tuan_hung_vu: #ECCV2024 .TL;DR: criterion for determining when to stop training in source-free domain adaptation. That is to avoid perf….
0
1
0
RT @hugues_va: Super happy to announce the release of TorchDR 0.1 : the Dimensionality Reduction (DR) library using @PyTorch. Joint work wi….
github.com
TorchDR - PyTorch Dimensionality Reduction. Contribute to TorchDR/TorchDR development by creating an account on GitHub.
0
31
0
RT @SashaMTL: "In this new generative-AI paradigm, uncertainty reigns over certainty, speculation dominates reality, science defers to fait….
theatlantic.com
Can artificial intelligence really enrich fossil-fuel companies and fight climate change at the same time? The tech giant says yes.
0
60
0
RT @RFlamary: Extremely happy to announce the (alpha) release 0.3.0 of SKADA Scikit Adaptation, a python toolbox for domain adaptation full….
0
27
0
RT @RFlamary: Python Optimal Transport (POT) 0.9.4 released : works with numpy 2+, new quantized and low rank Gromov-Wasserstein solvers,….
github.com
This new release contains several new features and bug fixes. Among the new features we have novel Quantized FGW solvers that can be used to speed up the computation of the FGW loss on large datase...
0
40
0
Congratulations @Eulerson314 for this award !.
With @nicolas_courty and @dcoeurjo, we just received the GĂĽnter Enderle Award for the best paper at @EurographicsC 2024, for our article "Non-Euclidean Sliced Optimal Transport Sampling"!. I should post a recording of my presentation very soon!. (photo by @boubek)
0
0
18
RT @_KarenHao: A big question looms over generative AI: what really is its impact on the environment? I spent months investigating a single….
0
664
0
RT @Eulerson314: I’m really glad to present my very first paper, “Non-Euclidean Sliced Optimal Transport Sampling” ! With @nicolas_courty a….
0
23
0
RT @RFlamary: We have an open position for tenure track assist. professor in data-centric engineering and applied ML at @Polytechnique. Com….
0
24
0
Thesis co-supervised with Lucas Drumetz and François Septier. His PhD document and slides will be soon put online.
0
1
6
Clément can be very proud of his significative contributions at the confluence of machine learning and optimal transport. Congratulations to the new doctor ! @UBS_universite @irisa_lab @IMTAtlantique LMBA.
0
2
8
Today was the PhD defense of the amazing Clément Bonet @Clement_Bonet_ for his work on 'Leveraging Optimal Transport via Projections on.Subspaces for Machine Learning Applications' with a dream jury @gabrielpeyre @GabrieleSteidl @FrnkNlsn @delonju @RFlamary @elmelis
4
17
206
RT @hugues_va: Excited to introduce SNEkhorn: our #NeurIPS2023 paper with @t_vayer, @RFlamary and @nicolas_courty. We propose a new affinit….
github.com
Contribute to PythonOT/SNEkhorn development by creating an account on GitHub.
0
11
0
RT @RFlamary: We are looking for a research engineer to work on domain adaptation and transfer learning @Polytechnique near Paris. Come wit….
0
50
0
RT @Clement_Bonet_: Happy to annouce that our work "Sliced-Wasserstein on Symmetric Positive Definite Matrices for M/EEG Signals" has been….
arxiv.org
When dealing with electro or magnetoencephalography records, many supervised prediction tasks are solved by working with covariance matrices to summarize the signals. Learning with these matrices...
0
10
0
RT @RFlamary: POT (Python Optimal Transport) 0.9.0 has been released on Pypi, Github and conda-forge. With many new features, 51 examples i….
0
75
0
5/ Overall, we show that the optimal transport cost is better, without changing the performances of the original NF in terms of log Likelihood. We also discuss in a disentanglement problem how enhancing the transport cost helps in better preserving the structure of data.
0
0
1
4/ We further constraint it to lie on a geodesic in the space of volume-preserving diffeomorphisms thanks to Euler’s equations. This leads to a better optimization problem.
1
0
3