AnyCalib: On-Manifold Learning for Model-Agnostic Single-View Camera Calibration
Camera calibration from a single perspective/distorted/edited image with closed-form solutions for a wide range of camera models.
I am a PhD student in the Robotics, Computer Vision, and Artificial Intelligence Group at the University of Zaragoza, supervised by Prof. Javier Civera. My research interests include solving 3D vision problems with Deep Learning and Geometry.
Camera calibration from a single perspective/distorted/edited image with closed-form solutions for a wide range of camera models.
An algorithm that avoids the 4-fold ambiguity of calibrated epipolar geometry and directly estimates the globally optimal relative pose from 2D correspondences.
Quantification of spatial (2D) uncertainty for the location of deep local features in an image and their inclusion in geometric tasks such as PnP and camera localization.
Continuous-time rigid object tracking using cumulative B-Splines on SE(3) with novel analytic Jacobians for speeding-up optimization.