Matrix-Free Shared Intrinsics Bundle Adjustment

Daniel Safari; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025, pp. 27017-27026

Research on accelerating bundle adjustment has focused on photo collections where each image is accompanied by its own set of camera parameters. However, real-world applications overwhelmingly call for shared intrinsics bundle adjustment (SI-BA) where camera parameters are shared across multiple images. Utilizing overlooked optimization opportunities specific to SI-BA, most notably matrix-free computation, we present a solver that is eight times faster than alternatives while consuming a tenth of the memory. Additionally, we examine factors contributing to BA instability under single-precision computation and propose mitigations.