Iterated Lifting for Robust Cost Optimization

Christopher Zach and Guillaume Bourmaud

Abstract

Optimization of latent model parameters using robust formulations usually creates a large number of local minima due to the quasi-convex shape of the underlying robust kernel. Lifting the robust kernel, i.e. embedding the problem into a higher-dimensional space, leads to significantly better local minima in a range of 3D computer vision problems (e.g. [10, 11, 12]). In this work we propose to iterate this lifting construction to obtain a more gradual lifting scheme for a given target kernel. Thus, a robust kernel is not directly lifted against the (non-robust) quadratic kernel, but initially against a different, less robust kernel. This process is iterated until the quadratic kernel is reached to allow utilization of efficient non-linear least-squares solvers.

Session

Posters

Files

PDF iconPaper (PDF)
PDF iconSupplementary (PDF)

DOI

10.5244/C.31.86
https://dx.doi.org/10.5244/C.31.86

Citation

Christopher Zach and Guillaume Bourmaud. Iterated Lifting for Robust Cost Optimization. In T.K. Kim, S. Zafeiriou, G. Brostow and K. Mikolajczyk, editors, Proceedings of the British Machine Vision Conference (BMVC), pages 86.1-86.11. BMVA Press, September 2017.

Bibtex

            @inproceedings{BMVC2017_86,
                title={Iterated Lifting for Robust Cost Optimization},
                author={Christopher Zach and Guillaume Bourmaud},
                year={2017},
                month={September},
                pages={86.1-86.11},
                articleno={86},
                numpages={11},
                booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
                publisher={BMVA Press},
                editor={Tae-Kyun Kim, Stefanos Zafeiriou, Gabriel Brostow and Krystian Mikolajczyk},
                doi={10.5244/C.31.86},
                isbn={1-901725-60-X},
                url={https://dx.doi.org/10.5244/C.31.86}
            }