# ETH Zurich's RaCo keypoint detector gets its rotation robustness from data augmentation, not from an equivariant architecture

The Apache 2.0-licensed model pairs a differentiable ranker for fixed keypoint budgets with a covariance estimator that reports spatial uncertainty in metric scale. It trains on single image crops without covisible pairs.

- Published: 2026-10-13T06:13:53.502Z
- Canonical: https://polylog.news/ai/2026-10-13/eth-zurich-s-raco-keypoint-detector-gets-its-rotation-robust
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [Polylog editors](https://polylog.news), [arXiv](https://arxiv.org/html/2602.15755), [arXiv (PDF)](https://arxiv.org/pdf/2602.15755)

The research channel AI with Papers circulated RaCo this week and described it as a "rotationally equivariant" keypoint detector, with code released under Apache 2.0 by the Computer Vision and Geometry group at ETH Zurich. The paper itself,…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-10-13/eth-zurich-s-raco-keypoint-detector-gets-its-rotation-robust (subscription information: https://polylog.news/pricing).