Romain Mueller

Research engineer, recovering physicist

[Romain Mueller @RomainMueller]

@lbertinetto EU accession speedrun? Is there a leaderboard?

[Romain Mueller @RomainMueller]

Do you want to work on perception/planning for self-driving cars? We have some intern positions available in the team next year! Apply or dms welcome :-)

[Jonathan Sadeghi @JonathanSadeghi]

Accepted for #AAAI23! 🎉 @RealAAAI

[Jonathan Sadeghi @JonathanSadeghi]

Paper accepted at @NeurIPSConf 2022 Workshop on Gaussian Processes, Spatiotemporal Modeling, and Decision-making Systems (GPSMDMS)! work with @RomainMueller and John Redford at @_FiveAI . Code available on github to reproduce results

[Jonathan Sadeghi @JonathanSadeghi]

New paper from @_FiveAI!🚗 https://t.co/uZ2In47Uhb
Can we efficiently calculate the reliability of a system when its performance function is partially defined? Usually we use Gaussian Process active learning algorithms but these don't work if performance is partially defined🔽1/4

[Romain Mueller @RomainMueller]

@paulg I think the correct formula is \integral curiosity(t) d t ;)

[Romain Mueller @RomainMueller]

@wightmanr @codinghorror Holy motors ♥️

[Jonathan Sadeghi @JonathanSadeghi]

New paper from @_FiveAI! https://t.co/F4B1Bc08Ru
How can we find images where 2D object detectors exhibit poor performance? It's an important problem for us to solve in autonomous driving, e.g. how can we find all the images where we miss out nearby pedestrians? 🔽 1/4

[Romain Mueller @RomainMueller]

Happy to share our work at @_FiveAI on perception verification! It is a new twist on finding hard images for object detectors at test time. Super important problem to help us solve autonomous driving https://t.co/Cu2lEFh92Y @puneetdokania @JonathanSadeghi @ewayers

[Romain Mueller @RomainMueller]

@BMVCconf any news about BMVC 2022? Deadline should be one month from now but there is still no program or definite deadline 😅

[Romain Mueller @RomainMueller]

@sinjax @jalayrac Super impressive results!

[Max Kleiman-Weiner @maxhkw]

The paper for PaLM (Google's new language model trained with 540B parameters on 780B tokens) breaks down the team structure for a mega scale machine learning project.
https://t.co/06j9BSvlZP

[Nicholas A. Lord @NicholasALord]

Hi. Do you care about adversarial attacks? Or not?

Either way, check out our #iclr2022 paper, "Attacking Deep Networks with Surrogate-Based Adversarial Black-Box Methods Is Easy".

Paper: https://t.co/vjswiiHxub

Video: https://t.co/6typikUzpT

More below. 👇 1/

[Nicholas A. Lord @NicholasALord]

Work done with @RomainMueller and @lbertinetto.

(Really simple) code will be available shortly at https://t.co/pVNaKH5U7h .

Use it on your own networks and tell us what you find. 🙏 6/6

[Romain Mueller @RomainMueller]

@phfaist Pas loin 😉 hope you have a good time

[Romain Mueller @RomainMueller]

@nathanbenaich @exscientiaAI @airstreet @GSK @AstraZeneca Congrats!

[Romain Mueller @RomainMueller]

@MedCrisis It's a really tough question... One would really hope it is not necessary, but if there is very low vaccine uptake among healthcare workers, is there a better solution?

[Romain Mueller @RomainMueller]

@remilouf Why not ubuntu? I use xubuntu because gnome no thx.

[Romain Mueller @RomainMueller]

@MedCrisis Yes! (Cambridge)