p-h duhemelle
@patrickduhemelle.bsky.social
Computer Science/ Machine Learning. ESR. Amour (non-reciproque) du service public.
Reposted by p-h duhemelle
Comme quoi, cette chanson des guignols de l'info est intemporelle.
January 3, 2026 at 5:26 PM
Comme quoi, cette chanson des guignols de l'info est intemporelle.
Il n’y a pas d’amour, il n’y a que des preuves d’amour.
Quand notre ministre de tutelle nous insulte devant la représentation nationale. "Bande de nuls" "complètement à la ramasse".
Nous reprocher des taux de réussite faible à Horizon Europe et ERC, quand manquent les moyens pour assurer nos missions de service public. Surtout changez rien!👌
Nous reprocher des taux de réussite faible à Horizon Europe et ERC, quand manquent les moyens pour assurer nos missions de service public. Surtout changez rien!👌
October 31, 2025 at 10:59 PM
Il n’y a pas d’amour, il n’y a que des preuves d’amour.
Reposted by p-h duhemelle
La macronie, cet ex toxique qui revient et te disant "nan mais j'ai changé"
October 5, 2025 at 7:41 PM
La macronie, cet ex toxique qui revient et te disant "nan mais j'ai changé"
Great and useful rant.
Where I rant on scale and current trends in research like a broken phonograph 😅
It's a great discussion because we all have different interpretation of what this means and what the consequences are.
Myself, I have absolutely no doubt that scaling works. If you have all the videos in the world and are able to train a model that can recall and merge any of them, then for sure...
Myself, I have absolutely no doubt that scaling works. If you have all the videos in the world and are able to train a model that can recall and merge any of them, then for sure...
October 3, 2025 at 11:02 AM
Great and useful rant.
Reposted by p-h duhemelle
This is not OK.
I don't submit often to NeurIPS, but I reviewed papers for this conference almost every year. As a reviewer, why would I spend time trying to give a fair opinion on papers if it's what happens in the end???
I don't submit often to NeurIPS, but I reviewed papers for this conference almost every year. As a reviewer, why would I spend time trying to give a fair opinion on papers if it's what happens in the end???
September 20, 2025 at 6:10 AM
This is not OK.
I don't submit often to NeurIPS, but I reviewed papers for this conference almost every year. As a reviewer, why would I spend time trying to give a fair opinion on papers if it's what happens in the end???
I don't submit often to NeurIPS, but I reviewed papers for this conference almost every year. As a reviewer, why would I spend time trying to give a fair opinion on papers if it's what happens in the end???
Reposted by p-h duhemelle
I hope that the ACs and SACs at @neuripsconf.bsky.social are being transparent and stating that the reason for rejecting an otherwise accepted paper was due to instructions from the PCs to reject it, rather than altering the reasons in the meta-review.
#NeurIPSRejects
#NeurIPSRejects
September 10, 2025 at 8:03 AM
I hope that the ACs and SACs at @neuripsconf.bsky.social are being transparent and stating that the reason for rejecting an otherwise accepted paper was due to instructions from the PCs to reject it, rather than altering the reasons in the meta-review.
#NeurIPSRejects
#NeurIPSRejects
Reposted by p-h duhemelle
One-day workshop on Diffusion models and Flow matching, October 24th at @ensdelyon.bsky.social
Registration and call for contributions (short talk and poster) are open at
gdr-iasis.cnrs.fr/reunions/mod...
Registration and call for contributions (short talk and poster) are open at
gdr-iasis.cnrs.fr/reunions/mod...
Modèles génératifs : diffusion, flow matching et leurs applications - GdR IASIS
Les modèles génératifs ont connu de récentes avancées spectaculaires, au point que leurs dernières versions sont désormais capables de produire des images et du texte synthétiques presque indiscernabl...
gdr-iasis.cnrs.fr
September 4, 2025 at 7:40 AM
One-day workshop on Diffusion models and Flow matching, October 24th at @ensdelyon.bsky.social
Registration and call for contributions (short talk and poster) are open at
gdr-iasis.cnrs.fr/reunions/mod...
Registration and call for contributions (short talk and poster) are open at
gdr-iasis.cnrs.fr/reunions/mod...
Reposted by p-h duhemelle
Distributional Reduction paper with H. Van Assel, @ncourty.bsky.social, T. Vayer , C. Vincent-Cuaz, and @pfrossard.bsky.social is accepted at TMLR. We show that both dimensionality reduction and clustering can be seen as minimizing an optimal transport loss 🧵1/5. openreview.net/forum?id=cll...
June 27, 2025 at 7:44 AM
Distributional Reduction paper with H. Van Assel, @ncourty.bsky.social, T. Vayer , C. Vincent-Cuaz, and @pfrossard.bsky.social is accepted at TMLR. We show that both dimensionality reduction and clustering can be seen as minimizing an optimal transport loss 🧵1/5. openreview.net/forum?id=cll...
Reposted by p-h duhemelle
Nothing says upcoming NeurIPS deadline like Overleaf being down.
May 14, 2025 at 6:23 AM
Nothing says upcoming NeurIPS deadline like Overleaf being down.
Reposted by p-h duhemelle
It's been 20 years and I think the new generation need to know about the SVM-KM toolbox. It was a Matlab open source SVM toolbox created in 2005 by @scanu.bsky.social, Yves Grandvalet, Vincent Guige, and Alain Rakotomamonjy 1/n github.com/rflamary/SVM...
February 14, 2025 at 4:11 PM
It's been 20 years and I think the new generation need to know about the SVM-KM toolbox. It was a Matlab open source SVM toolbox created in 2005 by @scanu.bsky.social, Yves Grandvalet, Vincent Guige, and Alain Rakotomamonjy 1/n github.com/rflamary/SVM...
on va y croire.
January 20, 2025 at 11:32 AM
on va y croire.
Rends le CIR.
Un (intéressant) article supplémentaire sur la gabegie du CIR (qui rend plus insupportable encore la misère de l'enseignement supérieur et de la recherche publics).
La face sombre du crédit d’impôt recherche
Première niche fiscale pour les entreprises, le dispositif de soutien à la recherche et au développement attire les convoitises et les pratiques frauduleuses.
www.lemonde.fr
January 10, 2025 at 8:11 AM
Rends le CIR.
strong stance of danah boyd on the new move of facebook.
The Ministry of Empowerment
Fuck you Facebook. That was the first thought I had when I woke up this morning. Followed by: What ministry is Mark Zuckerberg volunteering to manage for the dictators of the world? All I could think ...
made-not-found-by-danah-boyd.ghost.io
January 9, 2025 at 1:26 PM
strong stance of danah boyd on the new move of facebook.
38 ERC consolidator en France. 6 sont portés par des universités.
prior : si vous etes EC à l'université, peu de chance que vous soyez considéré comme excellent.
prior : si vous etes EC à l'université, peu de chance que vous soyez considéré comme excellent.
December 4, 2024 at 10:32 AM
38 ERC consolidator en France. 6 sont portés par des universités.
prior : si vous etes EC à l'université, peu de chance que vous soyez considéré comme excellent.
prior : si vous etes EC à l'université, peu de chance que vous soyez considéré comme excellent.
Blog post on flow matching with really cool illustrations.
Anne Gagneux, Ségolène Martin, @quentinbertrand.bsky.social Remi Emonet and I wrote a tutorial blog post on flow matching: dl.heeere.com/conditional-... with lots of illustrations and intuition!
We got this idea after their cool work on improving Plug and Play with FM: arxiv.org/abs/2410.02423
We got this idea after their cool work on improving Plug and Play with FM: arxiv.org/abs/2410.02423
November 27, 2024 at 9:28 AM
Blog post on flow matching with really cool illustrations.
This paper sounds pretty cool! fast video generation with high-order moments inside!
🍏 New preprint alert! 🍏
PoM: Efficient Image and Video Generation with the Polynomial Mixer
arxiv.org/abs/2411.12663
This is my latest "summer project" and it was so big I had to call in reinforcements (Thanks @nicolasdufour.bsky.social)
TL;DR Transformers are for boomers, welcome to the future
🧵👇
PoM: Efficient Image and Video Generation with the Polynomial Mixer
arxiv.org/abs/2411.12663
This is my latest "summer project" and it was so big I had to call in reinforcements (Thanks @nicolasdufour.bsky.social)
TL;DR Transformers are for boomers, welcome to the future
🧵👇
PoM: Efficient Image and Video Generation with the Polynomial Mixer
Diffusion models based on Multi-Head Attention (MHA) have become ubiquitous to generate high quality images and videos. However, encoding an image or a video as a sequence of patches results in costly...
arxiv.org
November 20, 2024 at 9:51 AM
This paper sounds pretty cool! fast video generation with high-order moments inside!