Assistant Professor @HeriotWattUni, Edinburgh. I like machine learning and imaging inverse problems.
edinburgh, uk
Joined December 2019
- Tweets46
- Following168
- Followers156
- Likes281
Dongdong Chen retweeted
DeepInverse has joined the PyTorch Ecosystem, making imaging with deep learning easier and more reproducible across research and industry.
DeepInverse is an open source framework for solving imaging inverse problems in medical imaging, computational photography, remote sensing, astronomical imaging, microscopy and more. DeepInverse makes imaging with deep learning easy and is developed by a passionate community of researchers, practitioners and engineers.
🔗 Learn more & explore how to get involved: hubs.la/Q03RKyHJ0
#PyTorch #OpenSourceAI #DeepInverse #Imaging #DeepLearning
Dongdong Chen retweeted
Not coming back from Marseille empty handed #DeepInverse @TachellaJulian @MatthieuTerris @ddongchen @minhhai712
Dongdong Chen retweeted
Toutes nos félicitations aux deux espoirs des prix science ouverte du logiciel libre de la recherche qui travaillent dans des laboratoires rattachés à @CNRSphysique : @TachellaJulian (DeepInverse) et Sébastien Weber (PyMoDAQ) 👏
Pour en savoir plus ⤵️
inp.cnrs.fr/fr/cnrsinfo/deep…
Dongdong Chen retweeted
🚀 deepinverse hackathon🚀
We just finished a couple of days working on deepinverse in the beautiful CIRM venue.
deepinv.github.io/deepinv/
Immensely grateful to the dream team of contributors that participated in the hackathon!
A thread 🧵
Dongdong Chen retweeted
📢New preprint 📢
UNSURE: Unknown Noise level Stein's Unbiased Risk Estimator
arxiv.org/abs/2409.01985v1
Optimal self-supervised losses (SURE, R2R) require exact knowledge of the noise distribution, and alternatives like Noise2Void are suboptimal.
There is middle ground🧵
Dongdong Chen retweeted
Tomorrow we are giving a deepinverse + computational imaging tutorial in Bologna, Italy, with fellow deepinverters @HuraultSamuel @MatthieuTerris @ddongchen
site.unibo.it/mathematical-m…
Dongdong Chen retweeted
Attending EUSIPCO'24 in Lyon?
You can attend a 3-hour tutorial on self-supervised learning for imaging, given by Mike Davies and myself.
eusipcolyon.sciencesconf.org…
A thread with some of the contents 🧵
"pip install deepinv" to solve your inverse problems🥳🥳
📢 DeepInverse v0.2 is out!
New features:
📏 Physics for blind inverse + calibration problems
🕸️ Advanced blurs
💣 Random physics generators
🤖New modular Trainer
🪂 Phase retrieval
🎉 Patch and 3D priors
by new contributors!
A🧵
deepinv.github.io/
Dongdong Chen retweeted
🤗You can now check out some deepinverse denoisers with this interactive huggingface demo
huggingface.co/spaces/deepin…
More demos are coming!
Dongdong Chen retweeted
💣The deepinverse library keeps growing!
Now you can train your model with the self-supervised loss 'recorrupted2recorrupted' by choosing the R2RLoss
Dongdong Chen retweeted
The group photo for DIPOpt at ENS Lyon, thanks everyone for coming and sharing ideas and research!
Update: Diffusion Posterior Sampling (DPS) is now integrated with deep inverse library (deepinv.github.io). Thanks @hyungjin_chung for the contribution!
'import deepinv as dinv' - one line of code to play with DPS.
🎁Diffusion Posterior Sampling (DPS) is now part of the deepinverse library.
Big thanks to @hyungjin_chung for the contribution!
The list of deepinv contributors keeps growing 🎉
Dongdong Chen retweeted
🚆On my way to #ICCV2023
Tomorrow I'll be giving an intro to the deepinverse library in the morning deepinv.github.io/deepinv/
📢🚨 @_LXAI is pleased to announce Dr. Julián Tachella @TachellaJulian (CNRS Lyon, France) as FEATURE SPEAKER of the @LatinXinCV Research Workshop at @ICCVConference 2023! 🇫🇷
For more information: buff.ly/46Qc6m7
Dongdong Chen retweeted
📢Happy to announce the "Deep learning, image analysis, inverse problems, and optimization" workshop on November 27-30 in Lyon, France.
perso.ens-lyon.fr/nelly.pust…
more details below 🧵
Do you need a library for inverse problems, computational imaging, or image reconstruction/restoration?
try 'pip install deepinv', then
'import deepinv' is all you need! 🥳
deepinv.github.io/
#deepinverse
Dongdong Chen retweeted
📢📢 Release of DeepInverse library 📢📢
After months of intense work, we are releasing the first stable version of DeepInverse deepinv.github.io/deepinv/, a PyTorch library for solving inverse problems with deep learning.
with @HuraultSamuel, @MatthieuTerris and @ddongchen
A 🧵
Dongdong Chen retweeted
Our paper has been accepted at the Journal of Machine Learning Research @JmlrOrg and is now available in final form in jmlr.org/papers/v24/22-0315.…
Looking forward to see these ideas applied in more imaging problems!
#imaging #deeplearning #unsupervised
Want to solve your imaging problem with deep learning but no ground-truth data for training?
Good news🥳! Learning from noisy and incomplete measurement data alone is possible:
"Sampling Theorems for Unsupervised Learning in Linear Inverse Problems" arxiv.org/abs/2203.12513
1/4
Dongdong Chen retweeted
Our review paper on "Imaging with Equivariant Deep Learning" is now available in final form in
ieeexplore.ieee.org/document… !
Our paper "Imaging with Equivariant Deep Learning" has been accepted at IEEE Signal Proc. Magazine🥳
with @ddongchen, Mike Davies, @caromitreka @MJEhrhardt and Ferdia Sherry.
We review how equivariance is used in deep networks for imaging. A thread🧵!
Dongdong Chen retweeted
Are you interested in inverse problems, learning from incomplete data or how to use deep learning for scientific discovery?
Come along to our poster #415 in the morning poster session today at #NeurIPS2022!
📢📢
Our paper "Unsupervised Learning From Incomplete Measurements for Inverse Problems" is accepted at
#NeurIPS2022!
We present
- theoretical analysis for learning from incomplete measurement data alone
- self-supervised loss which obtains SOTA results
A thread!
1/n
