Rotation Coordinate Descent for Fast Globally Optimal Rotation Averaging
This addresses the computational bottleneck in rotation averaging for computer vision and robotics, offering a faster global solution method.
The paper tackles the slow speed of semidefinite programming for globally optimal rotation averaging by introducing rotation coordinate descent (RCD), which directly updates valid rotations to avoid storing large matrices, achieving superior efficiency over state-of-the-art global methods.
Under mild conditions on the noise level of the measurements, rotation averaging satisfies strong duality, which enables global solutions to be obtained via semidefinite programming (SDP) relaxation. However, generic solvers for SDP are rather slow in practice, even on rotation averaging instances of moderate size, thus developing specialised algorithms is vital. In this paper, we present a fast algorithm that achieves global optimality called rotation coordinate descent (RCD). Unlike block coordinate descent (BCD) which solves SDP by updating the semidefinite matrix in a row-by-row fashion, RCD directly maintains and updates all valid rotations throughout the iterations. This obviates the need to store a large dense semidefinite matrix. We mathematically prove the convergence of our algorithm and empirically show its superior efficiency over state-of-the-art global methods on a variety of problem configurations. Maintaining valid rotations also facilitates incorporating local optimisation routines for further speed-ups. Moreover, our algorithm is simple to implement; see supplementary material for a demonstration program.