How DeepInverse Is Solving Imaging in Science and H...
summary
Claim that imaging is the bottleneck in computer vision and present a usable stack: DeepInverse PyTorch library ties physics models, differentiable ops and physics‑informed losses to deep nets and diffusion samplers. Demonstrations include accelerating brain MRI, reducing CT radiation and even imaging black holes via notebooks; microscope data scale hits 10,000³ voxels and clinical scanners process only tens of patients per day. Integrations with Hugging Face, a community hackathon, an ICLR paper and the Prix Science Ouverte 2024 show components working in practice.
AI-written summary · about
related talks
45mSolving Inverse Problems with Deep Learning
10mSeeing the World With AI: How Deep Learning Enables a New Era of Imaging Technologie
41mDeep Learning to Solve Challenging Problems
42mRunning Programs In Reverse for Deeper A.I.
34mArtificial intelligence to accelerate MRI scans - Philips & LUMC winners in the...
25mDeep Learning in Biomedicine: Challenges and Opportunities
29mReconstructing Unseen Dimensions in Biophotonics Through Python
51mDeep Learning - What is it and What It Can Do For You