Using superpixels instead of pixels has become a popular pre-processing step in computer vision. Currently, about fifteen oversegmentation methods have been proposed. The last evaluation, realized by Stutz et al. in 2015, concludes that the five more competitive algorithms achieve similar results. By introducing HSID, a new dataset, we point out unexpected difficulties encountered by state-of-the-art oversegmentation methods.
CITATION STYLE
Mathieu, B., Crouzil, A., & Puel, J. B. (2017). Oversegmentation methods: A new evaluation. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10255 LNCS, pp. 185–193). Springer Verlag. https://doi.org/10.1007/978-3-319-58838-4_21
Mendeley helps you to discover research relevant for your work.