Abstract
A number of computer vision deep regression approaches report improved results when adding a classification loss to the regression loss. Here, we explore why this is useful in practice and when it is beneficial. To do so, we start from precisely controlled dataset variations and data samplings and find that the effect of adding a classification loss is the most pronounced for regression with imbalanced data. We explain these empirical findings by formalizing the relation between the balanced and imbalanced regression losses. Finally, we show that our findings hold on two real imbalanced image datasets for depth estimation (NYUD2-DIR), and age estimation (IMDB-WIKI-DIR), and on the problem of imbalanced video progress prediction (Breakfast). Our main takeaway is: for a regression task, if the data sampling is imbalanced, then add a classification loss.
Original language | English |
---|---|
Title of host publication | Proceedings of the 2023 IEEE/CVF International Conference on Computer Vision (ICCV) |
Editors | Cristina Ceballos |
Place of Publication | Piscataway |
Publisher | IEEE |
Pages | 19915-19924 |
Number of pages | 10 |
ISBN (Electronic) | 979-8-3503-0718-4 |
ISBN (Print) | 979-8-3503-0719-1 |
DOIs | |
Publication status | Published - 2023 |
Event | 2023 IEEE/CVF International Conference on Computer Vision (ICCV) - Paris, France Duration: 1 Oct 2023 → 6 Oct 2023 |
Publication series
Name | Proceedings of the IEEE International Conference on Computer Vision |
---|---|
ISSN (Print) | 1550-5499 |
Conference
Conference | 2023 IEEE/CVF International Conference on Computer Vision (ICCV) |
---|---|
Country/Territory | France |
City | Paris |
Period | 1/10/23 → 6/10/23 |
Bibliographical note
Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-careOtherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.