Vaibhav Agrawal

@TacoCohen good old lua torch and backdrop by ~hand days.

@TheMartinsLab Congratulations Andre!

Oh wow! Thank you. 🥸

Can we detect type2 diabetes from full-body magnetic resonance imaging (MRI)? We used deep learning & data from 2400 MRIs to identify patients with (pre-)diabetes, incl. subgroups with high future diabetes risk. @diabresearch @uni_tue @MPI_IS @wagrob32 insight.jci.org/articles/view/…

#Radiomics risk scores were used to predict progression-free survival in patients with #glioblastoma doi.org/10.1148/ryai.2… @Ruchika__Verma @Pallavi_Tiwari @CWRUBME #BrainTumor #ML #MachineLearning

Statistical distances between densities of an exponential family often admit closed-form formula (Sharma-Mittal/Renyi/Tsallis/Shannon relative entropies/divergences). Optimal transport distances vs information-geometric divergences. Projective divergences tinyurl.com/y67hmoxs

Hot off the press: We have developed a self-supervised learning method that is much better for pre-training than ImageNet--reduces labeling needs by an order of magnitude for medical imaging applications: @stanfordAIMI @Radiology_AI

Some recently proposed distances with their rationales

Causalworld:A Robotic Manipulation Benchmark For Causal Structure And Transfer Learning arxiv.org/abs/2010.04296 sites.google.com/view/causal-wo…

Excited to share our recent work on predicting cortical metrics across a range of brain development stages and pathologies from MRI using #MachineLearning #DeepLearning #AI. Presented at @MLCNworkshop @MICCAI2020 #MLCN2020 #MICCAI2020 arxiv.org/abs/2010.01669

HEALTHCARE: By accessing the rich data from the Canadian #healthcare system, Pineau believes we can predict outcomes such as the likelihood of a disease recurring. Full story via @CIFAR_News: cifar.ca/cifarnews/2020… #Canada #artificalintelligence

Even animals can receive virtual health care with the help of #computervision! spr.ly/6017GLgCP

Our friends @ml4science are hosting a virtual workshop on the philosophy of #MedicalAI this week. Check out the event schedule here: #c1168441" target="_blank" rel="nofollow noopener">uni-tuebingen.de/en/165313#c116… #CyberValley #AI #MachineLearning @uni_tue

New paper: we characterize optimal representations for supervised learning, and show how to ~learn them! Our framework gives 1) a regularizer and 2) a predictor of generalization in DL. arxiv.org/abs/2009.12789 (NeurIPS spotlight) with @douwekiela @davidjschwab @Rama_vedantam 1/7

A first-time cooperation between @metin_sitti and his team and @mpifkf resulted in a @PNASNews publication showing a solar-battery effect that enables a new light-driven organic #microswimmer to operate in the dark. Find out more: is.mpg.de/news/solar-bat…

RigL is a new algorithm for training sparse neural networks. Instead of pruning a pre-existing dense network, it dynamically builds one during training without sacrificing accuracy relative to traditional approaches. Learn how it’s done at goo.gle/2ZJryiU

Natural gradient uses *steepest descent* but may leave the manifold. Riemannian gradient always stay on the manifold (but exponential map difficult to calculate). Natural gradient= approximation of the Riemannian gradient using a simple rectraction. tinyurl.com/y43bu2zd

1/4) We are pleased to announce TriFinger: an open-source robot for learning dexterity sites.google.com/view/trifinger

TriFinger: An Open-Source Robot for Learning Dexterity arxiv.org/abs/2008.03596

We are pleased to announce the start of the real robot challenge! Participate in the simulation phase now and qualify for hundreds of real-robot hours! Get started here: real-robot-challenge.com/simulation_pha…

🤖Calling all #deeplearning / #machinelearning / #reinforcementlearning or #AI scientists - no matter where in the world - to take part in a challenge that aims to advance the state of the art in #robotic manipulation: real-robot-challenge.com @robo_challenge #RoboChallenge🤖