Michal Drozdzal
@michal_drozdzal
Followers
156
Following
302
Media
7
Statuses
171
Joined January 2015
Happy to announce that our paper “On learning adaptive acquisition policies for undersampled multi-coil MRI reconstruction” has been accepted to #MIDL2022! Joint work with the amazing @luisenp, @mattmucklm, @adri_romsor, and @michal_drozdzal as part of the FAIR internship. 🧵
1
6
24
It will only cost 2% extra investment of our annual GDP to stay below 1.5 degrees climate change. That’s around 1.7 trillion a year. We spend 1/3 of that on direct subsidies to fossil fuels. Tax Carbon emissions and that to invest in sustainability!
4
28
192
We're delighted to share @uofg student @BorisAKnyazev's work during a 1+ year collaboration with @facebookai. Special thanks to wonderful collaborators @adri_romsor and @michal_drozdzal and @VectorInst for code review. Seems unbelievable?
github.com
Code for Parameter Prediction for Unseen Deep Architectures (NeurIPS 2021) - facebookresearch/ppuda
Do we still need SGD/Adam to train neural networks? Based on our #NeurIPS2021 paper, we are one step closer to replacing hand-designed optimizers with a single meta-model. Our meta-model can predict parameters for almost any neural network in just one forward pass. (1/n)
0
10
45
We are looking to hire FAIR research interns (current PhD students) in 2022. Previous works led by students: https://t.co/si1C0dkJ8J,
https://t.co/gPxAKeruef,
https://t.co/Bt4qjsXt4B,
https://t.co/mWpb7wsPk5,
https://t.co/7OSEh1l9RR. DM me or @michal_drozdzal with bio and CV.
4
24
133
Do we still need SGD/Adam to train neural networks? Based on our #NeurIPS2021 paper, we are one step closer to replacing hand-designed optimizers with a single meta-model. Our meta-model can predict parameters for almost any neural network in just one forward pass. (1/n)
29
303
2K
We’ve built a new image generation model called Instance-Conditioned GAN. Unlike traditional GANs, IC-GANs can create high-quality images of entirely unfamiliar objects and scenes that aren’t part of the model’s training data. Code+ details:
ai.meta.com
We’re sharing details on how we built a new image generation model called Instance-Conditioned GAN. Unlike traditional GANs, IC-GANs can create high-quality images of entirely unfamiliar objects and...
5
125
576
I'm very excited to have joined @luisenp, @adri_romsor and @michal_drozdzal at FAIR Montreal today as a summer research intern!
0
2
28
Science and Engineering of Deep Learning is happening tomorrow @iclr_conf! We are very excited about the program, please check it out at:
sites.google.com
SEDL @ FAccT 2022 Join us here! 21-24 June 2022
1
7
18
🎤 In her #WTGC2021 talk, @adri_romsor will discuss leveraging the complementarity of different data modalities, and actively acquiring information from the same data modality. 🎫 Reserve Your Ticket: https://t.co/OezdBFAS78
#womenintech #machinelearning
0
8
10
(1/n) Anyone who's needed an MRI understands how uncomfortable and expensive it is. Many others needed one but couldn't get one because these limitations. This is my I'm excited about further work on accelerating MRI's via AI: https://t.co/tQhpUYinsw
ai.meta.com
The future of healthcare is personal. Today we’re publishing research on active MRI acquisition, which uses reinforcement learning to adjust AI-accelerated scans on a
5
26
134
Really happy to share our work on using reinforcement learning for accelerating MRI acquisition, which we presented at @MICCAI2020 today (joint work with amazing collaborators @sumanabasu21 @adri_romsor @RCalandra @michal_drozdzal). 1/7
The future of health care is personal. Today, we’re publishing research on active MRI acquisition, which uses reinforcement learning to adjust AI-accelerated scans on a per-patient basis. This work complements our previous #fastMRI research w/@NYULangone. https://t.co/UEpBw6ufoT
1
8
25
The future of health care is personal. Today, we’re publishing research on active MRI acquisition, which uses reinforcement learning to adjust AI-accelerated scans on a per-patient basis. This work complements our previous #fastMRI research w/@NYULangone. https://t.co/UEpBw6ufoT
3
33
107
Checkout this blogpost about about work on histopathology images - some of that is related to what I was working on during my internship with @michal_drozdzal at @facebookai ( https://t.co/oWnK4EhFoR)
With collaborators at Radboud University Medical Centre and Erasmus Medical Center, we propose a new gigapixel histopathology image classification pipeline based on CNNs. https://t.co/zvSwLdqoIO
0
2
20
With collaborators at Radboud University Medical Centre and Erasmus Medical Center, we propose a new gigapixel histopathology image classification pipeline based on CNNs. https://t.co/zvSwLdqoIO
5
56
142
The first-ever fastMRI challenge is here! Based on a research project by Facebook AI and NYU Langone Health, this challenge aims to produce diagnostic-quality images in less time. Winning teams will be invited to present at a NeurIPS 2019 workshop. https://t.co/wZbJ9VhMhQ
5
151
427
@ZizhaoZhang @michal_drozdzal and Matt presenting our work on active acquisition for MRI reconstruction now at #CVPR2019. Stop by poster 207!
1
2
1
Barcelona is trying to show the world that there's a way out for loud, polluted, congested, warming cities. Around the world, hundreds of cities are looking on, waiting to see if it’s true.
vox.com
A few US cities, like Portland, Oregon, might have the infrastructure and civic will to follow in Barcelona’s footsteps.
0
21
56
More mature viewpoint: In a free, dynamic, & tradeoff-laden society, problems are inevitable, & no one is brilliant & omniscient enough to know the perfect solution. Better to seek new ideas, evidence for best tradeoffs than to blame (rival) "elites."
nytimes.com
France is pivotal in holding the European Union together.
19
45
197
Our @nyulangone Radiology Chair, Dr. Michael Recht, just announced public release of the largest-ever repository of raw MRI data, to enable #AI research, as part of our fastMRI collaboration. Use it in (and for) good health! https://t.co/m5EpcFEzil
@RSNA #RSNA18 @cai2r
2
90
207
Florian Knoll (NYU) and @MichalDrozdzal (@facebookai) announcing FastMRI at ISMRM Machine Learning Workshop
2
2
6