Rethinking Long-Tailed Visual Recognition with Dynamic Probability Smoothing and Frequency Weighted Focusing
Document Type
Conference Item
Publication Date
1-1-2023
Abstract
Deep learning models trained on long-tailed (LT) datasets often exhibit bias towards head classes with high frequency. This paper highlights the limitations of existing solutions that combine class- and instance-level re-weighting loss in a naive manner. Specifically, we demonstrate that such solutions result in overfitting the training set, significantly impacting the rare classes. To address this issue, we propose a novel loss function that dynamically reduces the influence of outliers and assigns class-dependent focusing parameters. We also introduce a new long-tailed dataset, ICText-LT, featuring various image qualities and greater realism than artificially sampled datasets. Our method has proven effective, outperforming existing methods through superior quantitative results on CIFAR-LT, Tiny ImageNet-LT, and our new ICText-LT datasets. The source code and new dataset are available at https://github.com/nwjun/FFDS-Loss.
Keywords
Long-tailed Classification, Weighted-loss
Divisions
fsktm
Publisher
IEEE
Publisher Location
345 E 47TH ST, NEW YORK, NY 10017 USA
Event Title
2023 IEEE International Conference on Image Processing (ICIP)
Event Location
Kuala Lumpur, Malaysia
Event Dates
8-11 October 2023
Event Type
conference
Additional Information
30th IEEE International Conference on Image Processing (ICIP), Kuala Lumpur, MALAYSIA, OCT 08-11, 2023