Florence Forbes retweeted
Congrats to Dr Yuchen Bai who successfully defended his PhD on AI for LiDAR Point Cloud Analysis! Yuchen was co-advised at @InriaStatify by @FForbes70 and Jean Baptiste Durand.
na1an.github.io/
Florence Forbes retweeted
🎓Congratulations to Dr. @Thomas__Coudert for successfully defending his PhD!
🌟 He presented his impressive work on "#MRI Fingerprint and Artificial Intelligence for #stroke emergencies".
#MRF #AI
Florence Forbes retweeted
Very happy to see our work on fast/accurate/lightweight SBI published in Transactions on Machine Learning Research
Fast, accurate and lightweight sequential simulation-based inference using Gaussian locally linea...
Henrik Häggström, Pedro L. C. Rodrigues, Geoffroy Oudoumanessah, Florence Forbes, Umberto Picchini.
Action editor: Alp Kucukelbir.
openreview.net/forum?id=Q0nz…
#
Florence Forbes retweeted
BCF #BestCollaboratorsForever @FForbes70 @TresBienHien between 🇦🇺 @latrobe and 🇫🇷 @inria_grenoble
Florence Forbes retweeted
Great @ISBA_events session on Bayesian computations. Healthy mix of speakers @FForbes70 @tmoellenhoff @TresBienHien newcomers to ISBA & nice handwritten discussion by @RousseauJudith who was a newcomer in the previous century :)
Florence Forbes retweeted
Limitations of the work: see the Disclaimer in the blogpost wp.me/p7WuDT-Yy There are some design choices to be made when comparing methods, and setups are never "universal". Things can be improved and this is just a start in an alternative direction of SBI.
Florence Forbes retweeted
Special kudos to Henrik (chalmers.se/en/persons/henha…), as this work started as his master thesis project! 👏 8/n
Florence Forbes retweeted
new paper on simulation-based inference using Gaussian mixtures to learn likelihood and posterior, without neural networks 😱😱😱
Our SeMPLE (Sequential Mixture Posterior and Likelihood Estimation) provides excellent inference and excels when the posterior is multimodal. 1/n
Florence Forbes retweeted
Summary blogpost: wp.me/p7WuDT-Yy
Paper: arxiv.org/abs/2403.07454
Code: github.com/henhagg/semple
This was work with the excellent Henrik Häggström, @plc_rodrigues , Geoffroy Oudoumanessah, @FForbes70 A real pleasure to work with you all! 7/n
Florence Forbes retweeted
A wonderful conclusion to these 3 years of thesis. 🎉
Thank you kindly to the jury members for the exciting discussions about the challenges of AI in clinical routine 🏥
@NinonBurgos @marazul @c_f_baumgartner @julienmairal @fran_galassi Olivier François & Pierrick Coupé
🎓 Congratulations to Dr. Benjamin Lambert for brilliantly defending his PhD!
💡 His research on 'Uncertainty Quantification in Deep Learning-based Medical Image Segmentation' was presented with impressive clarity and pedagogy. 🧠🤖
#IA #MRI #DeepLearning
Florence Forbes retweeted
Thrilled to announce that our paper "Robust Conformal Volume Estimation in 3D Medical Images" has been early accepted at #MICCAI2024🎉
A beautiful conclusion to my PhD journey, before my defense this Thursday.
Stay tuned for more info!
@nifm_gin @InriaStatify @pixylmedical
Florence Forbes retweeted
📢PhD defense on 16/05 at 10am! 🎓
🤖How can we guarantee the safe deployment of deep learning models in medical imaging?
🗣️Benjamin Lambert dives into this issue with his thesis: "#Uncertainty Quantification in Deep Learning-based Medical Image segmentation".
#AI #MRI #medicalAI
Florence Forbes retweeted
🔍 @TChristenGIN a présenté ses travaux en session orale à l' #IABM2024.
🌟 La communauté des chercheurs en Intelligence Artificielle en Imagerie BioMédicale a découvert les dernières avancées en #IRM ultra-rapides par IRM Fingerprint. 🧠
#MRF #MRI
Florence Forbes retweeted
🌟 @MichelDojat a animé la table ronde autour des enjeux éthiques de l'Intelligence artificielle durant l' #IABM2024.
🗣️ Ce temps d'échange a permis d'aborder différents aspects spécifiques à l'Imagerie BioMédicale rencontrés par la communauté de chercheurs.
Florence Forbes retweeted
Interested in trustworthy AI for medical applications? see sciencedirect.com/science/ar…
Florence Forbes retweeted
Just had my mind blown by @RobinJRyder's seminar on Bayesian inference in Historical Linguistics, partly based on a paper just accepted in Science @sciencemagazine that was still under embargo during the talk 😅🤯📚
Thanks for visiting Robin!
New paper out in Science @sciencemagazine!
Computational phylogenetics reveal histories of sign languages ⬇️
science.org/doi/10.1126/scie…
Florence Forbes retweeted
Two workshop papers on improving deep ensembles will be presented at @NeurIPSConf WANT@NeurIPS by @kostaspitas_
openreview.net/pdf?id=VN4tOg…
openreview.net/pdf?id=oYztmX…
Very happy to announce that 2 of my submissions on improving deep ensembles, together with @JulyanArbel and @MichaelArbel have been accepted at the WANT(Workshop on Advanced Neural Network Training) @NeurIPSConf !
Florence Forbes retweeted
Check out this review paper on Bayesian neural networks by @InriaStatify members and coauthors.
Interested in more details? Check out (and hopefully cite) our review paper which joint work with @JulyanArbel @iamvladimirova @vincefort arxiv.org/abs/2309.16314
