Anima's team unveils CTO, first neural operator for resolution-agnostic sparse-view CT
AnimaAnandkumar · x · 2026-09-13
Anima Anandkumar's group published an ECCV paper introducing CTO (Computed Tomography neural Operator), the first neural operator framework for sparse-view CT reconstruction.
- Sparse-view CT cuts radiation dose and scan time by taking fewer projections, but reconstruction becomes ill-posed and requires a learned prior
- Existing deep models are tied to one subsampling rate, while clinical protocols vary by organ and diagnostic purpose, forcing a separate model per rate
- CTO learns a mapping between function spaces: one model ingests sinograms at any subsampling rate and outputs high-quality reconstructions with no retraining
- On average, CTO beats traditional unrolled CNN variational methods
More from Research
- AIR: Effective-Learning-Rate Scheduler Beats Cosine on 128B-Token MoE Pretraining — stochasticchasm · 2026-10-30
- CMU's Gym-Anything Auto-Builds 200+ Agent Environments from Any Software, Selected by GDP — wellecks · 2026-10-30
- NVIDIA researcher: train robot policies with privileged sim info, improving Pi 0.5 — wellecks · 2026-10-30
- Leveraging Simulator Privileged Information: Three Methods to Improve Robot Models — wellecks · 2026-10-30
- CMU Auto-Builds 25 Robot Simulations with Agents; Frontier Model Solves 0 of 25 Test Tasks — wellecks · 2026-10-30
- DeepMind's f_charton shares his take on the current state of AI and mathematics in Sorbonne talk — f_charton · 2026-10-30