Criteria Sliders: Learning Continuous Database Criteria via Interactive Ranking

James Tompkin, Kwang In Kim, Hanspeter Pfister and Christian Theobalt

Abstract

Large databases are often organized by hand-labeled metadata—or criteria—which are expensive to collect. We can use unsupervised learning to model database variation, but these models are often high dimensional, complex to parameterize, or require expert knowledge. We learn low-dimensional continuous criteria via interactive ranking, so that the novice user need only describe the relative ordering of examples. This is formed as semi-supervised label propagation in which we maximize the information gained from a limited number of examples. Further, we actively suggest data points to the user to rank in a more informative way than existing work. Our efficient approach allows users to interactively organize thousands of data points along 1D and 2D continuous sliders.

Session

Spotlights

Files

Paper (PDF)
Video (MP4)

DOI

10.5244/C.31.105
https://dx.doi.org/10.5244/C.31.105

Citation

Bibtex

            @inproceedings{BMVC2017_105,
                title={Criteria Sliders: Learning Continuous Database Criteria via Interactive Ranking},
                author={James Tompkin, Kwang In Kim, Hanspeter Pfister and Christian Theobalt},
                year={2017},
                month={September},
                pages={105.1-105.12},
                articleno={105},
                numpages={12},
                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.31.105},
                isbn={1-901725-60-X},
                url={https://dx.doi.org/10.5244/C.31.105}
            }