5 citations · 12 across the 12 of their papers we have counts for
3 papers · 1 filter
Implicit Learning of Scene Geometry from Poses for Global Localization
Mohammad Altillawi, Shile Li, Sai Manoj Prakhya +2
Global visual localization estimates the absolute pose of a camera using a single image, in a previously mapped area. Obtaining the pose from a single image enables many robotics a…
Global Localization: Utilizing Relative Spatio-Temporal Geometric Constraints from Adjacent and Distant Cameras
Mohammad Altillawi, Zador Pataki, Shile Li +1
Re-localizing a camera from a single image in a previously mapped area is vital for many computer vision applications in robotics and augmented/virtual reality. In this work, we ad…
Long-Term Invariant Local Features via Implicit Cross-Domain Correspondences
Zador Pataki, Mohammad Altillawi, Menelaos Kanakis +5
Modern learning-based visual feature extraction networks perform well in intra-domain localization, however, their performance significantly declines when image pairs are captured…