🚨JOB ALERT🚨 Join me and the medical imaging team at Microsoft Research Health Futures as a Multi Modal Deep Learning Researcher. Apply now: https://jobs.careers.microsoft.com/global/en/job/1581473/Microsoft-Research-Residency-Program-%E2%80%93-Multi-Modal-Deep-Learning-Researcher. We’re looking for someone interested in multi-modal learning and medical imaging, working on text-to-image, image-to-text, and contrastive learning methods. To get an idea of our research agenda check out our latest work: https://arxiv.org/abs/2301.04558.
Maximilian Ilse
- 170 Followers
- 150 Following
- 8 Posts
Senior Researcher @ Health Futures - Microsoft Research. he/him.
We’ve developed a new method for jointly training with X-rays and radiology reports, published a new benchmark dataset, and performed extensive evaluations. Plus, we’ve open-sourced all the code (https://github.com/microsoft/hi-ml/tree/main/hi-ml-multimodal) and weights (https://huggingface.co/microsoft/BiomedVLP-BioViL-T).
I will give a talk at the Causality Discussion group tomorrow, 16:30 CEST! Everyone can join online for free! http://discuss.causality.link I will be talking about how the effectiveness of data augmentation can be explained by causality!
Owen Seymour@jmlipman @wendyfyx
@afoncubiertar
@kiran
@gxr
@kaushalya
@seabbs
@DrVeronikaCH
@raghavian
@AtoAndyKing
@victorvicpal
@Mo
I've added all of you so far. If you don't want to be included you should be able to remove yourself (and my apologies, but I wanted to get this community kicked off somehow)
@afoncubiertar
@kiran
@gxr
@kaushalya
@seabbs
@DrVeronikaCH
@raghavian
@AtoAndyKing
@victorvicpal
@Mo
I've added all of you so far. If you don't want to be included you should be able to remove yourself (and my apologies, but I wanted to get this community kicked off somehow)
Also, please use #MedicalAI so we can all find each other! (credit to @jmlipman and @wendyfyx for the idea)
Jonas KöhlerFor those who don't know it yet: https://moa.party is a super useful crossposting tool, that allows you to automatically cross-post toots to the bird app.
Using # you can also steer which toots to share.
Helpful to still reach your 🐦 audience with a spoiler while redirecting them to the better alternative for the full fun 🤓
My Phd thesis: Invariance in Deep Representations is finally online! https://dare.uva.nl/search?identifier=1784763c-d663-4626-bae3-d3c3735bcc13
The thesis combines equivariant/invariant neural network architectures, disentanglement (via generative models), causality, and data augmentation in one framework.
I'm a senior researcher at Microsoft Research Health futures in Amsterdam/Cambridge.
My research interests include invariance/equivariance, causality and representation learning, especially the interplay of these 3 things. All applied to medical imaging.