Class-balanced loss based on effective number of samples
Publikation: Bidrag til tidsskrift › Konferenceartikel › Forskning › fagfællebedømt
With the rapid increase of large-scale, real-world datasets, it becomes critical to address the problem of long-tailed data distribution (i.e., a few classes account for most of the data, while most classes are under-represented). Existing solutions typically adopt class re-balancing strategies such as re-sampling and re-weighting based on the number of observations for each class. In this work, we argue that as the number of samples increases, the additional benefit of a newly added data point will diminish. We introduce a novel theoretical framework to measure data overlap by associating with each sample a small neighboring region rather than a single point. The effective number of samples is defined as the volume of samples and can be calculated by a simple formula (1-beta {n})/(1-beta), where n is the number of samples and beta in [0,1) is a hyperparameter. We design a re-weighting scheme that uses the effective number of samples for each class to re-balance the loss, thereby yielding a class-balanced loss. Comprehensive experiments are conducted on artificially induced long-tailed CIFAR datasets and large-scale datasets including ImageNet and iNaturalist. Our results show that when trained with the proposed class-balanced loss, the network is able to achieve significant performance gains on long-tailed datasets.
Originalsprog | Engelsk |
---|---|
Tidsskrift | Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition |
Sider (fra-til) | 9260-9269 |
Antal sider | 10 |
ISSN | 1063-6919 |
DOI | |
Status | Udgivet - jun. 2019 |
Eksternt udgivet | Ja |
Begivenhed | 32nd IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2019 - Long Beach, USA Varighed: 16 jun. 2019 → 20 jun. 2019 |
Konference
Konference | 32nd IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2019 |
---|---|
Land | USA |
By | Long Beach |
Periode | 16/06/2019 → 20/06/2019 |
Bibliografisk note
Publisher Copyright:
© 2019 IEEE.
ID: 301824490