Michael P. Sheehan

PhD student researching Statistical and Compressive Learning at the University of Edinburgh.

[Sotirios (Sotos) Tsaftaris @STsaftaris]

Please re-tweet: PhD position in causality, representation learning & applications in healthcare. Fully funded international fees too (by @CanonMedicalEDI @RAEngNews). Co-supervised by @alisonQoneil. Join @SnchzPedro_ @Xiaoliu33745946 and others in VIOS.
https://t.co/OM2DcsrY0L

[Low-rank Jack (go mathstodon) @jacquesdurden]

An open faculty position in "Computational Mathematics and Data Science" is announced in the ICTEAM Institute of the University of Louvain, with an application deadline on November 14, 2022. https://t.co/KFEvzAmOY3

[Dongdong Chen @ddongchen]

🥳Our paper "Unsupervised Learning From Incomplete Measurements for Inverse Problems" was accepted to #NeurIPS2022. We present necessary and sufficient conditions for learning the signal model + a novel self-supervised Multi-Operator Imaging (MOI) method.
https://t.co/QBvfznnYut

[Sotirios (Sotos) Tsaftaris @STsaftaris]

Please retweet widely. Looking for a postdoc in representation learning. We want to learn good representations with (causal) generative models with rare/imbalanced data. You will work with bright minds, on a prestigious grant and interact with industry. DM me for more info.

[Julián Tachella @TachellaJulian]

Which dataset would you choose for training the reconstruction net of an imaging system?

1. Pairs of images+measurements
2. Just all raw measurements

Spoiler alert: the obvious answer is incorrect - a thread 🧵

[Pedro P. Sanchez @SnchzPedro_]

Our paper on disentangled representations is available online! Go check it out!

[Julián Tachella @TachellaJulian]

Attending #CVPR2022? Interested in imaging inverse problems? @ddongchen will be presenting our work in fully unsupervised learning for imaging inverse problems at Oral 2.1.3: Low-Level Vision (Great Hall B-C)

[School of Engineering @SchoolOfEng_UoE]

Our researchers scooped Best Student Paper at @ieeeICASSP for a #lidar breakthrough that could bring self-driving cars closer to market.

Dr Mikey Sheehan, Prof Mike Davies & Dr Julián Tachella are working with @IbeoAutomotive to commercialise the tech.

https://t.co/CtsUBbhHOu

[Michael P. Sheehan @mpsheehan1995]

@jacquesdurden @schellekensv Congrats @schellekensv, I took great inspiration from this thesis!

[IEEE ICASSP @ieeeICASSP]

Best Student Paper Awards! Congratulations to all the award winning authors!! (2/2)
#ICASSP2022

[Michael P. Sheehan @mpsheehan1995]

I'm very proud to say our paper "Sketched RT3D: how to reconstruct billions of photons per second" received the best student paper at @ieeeICASSP this year. Thanks to my co-authors @TachellaJulian and Mike Davies for their great work. Check it out here - https://t.co/GwP5zuNrap

[Michael P. Sheehan @mpsheehan1995]

Check out our recent work at @ieeeICASSP on reconstructing billions of photons a second using sketched lidar methods

[Julián Tachella @TachellaJulian]

Attending @ieeeICASSP 2022?

If you are interested in single-photon lidar and/or sketching methods, please join me in the
CI-3: Computational Photography session tomorrow, 10 May, 17:00 - 17:45 France Time (UTC +2)

Joint work with @mpsheehan1995 and Mike Davies.

[Julián Tachella @TachellaJulian]

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" https://t.co/n6IGQzFmMa
1/4

[Pedro P. Sanchez @SnchzPedro_]

(1/3) I am happy to share that our paper "Diffusion Causal Models for Counterfactual Estimation" will be presented next week at @CLeaR_2022. Check poster session 1 (11/04 - 4pm) or our 90s video! Huge thanks to @STsaftaris and the rest of the team at https://t.co/ZkaE9rqdfs!

[Dongdong Chen @ddongchen]

Check our latest theory paper (with @TachellaJulian and Mike Davies) on unsupervised learning in inverse problems. This is the first-ever work presents necessary and sufficient sampling conditions for learning the signal model from partial measurements!
https://t.co/At1cX2jJWj

[Michael P. Sheehan @mpsheehan1995]

@ddongchen @TachellaJulian Congrats guys!

[Dongdong Chen @ddongchen]

Robust Equivariant Imaging (REI) got in #CVPR2022! Joint work with @TachellaJulian and Mike Davies.🍻

Paper: https://t.co/cI7xWnxZMG
Code: https://t.co/sdxMyuZ60M

As a fully unsupervised framework, REI can learn the reconstruction function from noisy and partial measurements.

[Peyman Milanfar @docmilanfar]

This isn’t a scene from Tenet. It’s a great example of critical-rate sampling.

Temporal aliasing occurs when the scene changes faster than camera frame rate.

‘Critical sampling’ is when the scene changes exactly as fast as the camera frame rate.

https://t.co/LHwQD9NKLE

[Sotirios (Sotos) Tsaftaris @STsaftaris]

Please spread widely: PhD position in causality, representation learning and applications in healthcare. (funded by @CanonMedicalEDI @RAEngNews). Co-supervised by @alisonQoneil. Come and join @SnchzPedro_ @Xiaoliu33745946 @jacenkow and the VIOS team.
https://t.co/OM2Dcsrqbd

[Stat.ML Papers @StatMLPapers]

Mean Nystr\"om Embeddings for Adaptive Compressive Learning. (arXiv:2110.10996v1 [https://t.co/zjV5HgYw5a]) https://t.co/WzEGBfdKmx