English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In this paper, our goal is to design a simple learning paradigm for long-tailvisual recognition, which not only improves the robustness of the featureextractor but also alleviates the bias of the classifier towards head classeswhile reducing the training skills and overhead. We propose an efficientone-stage training strategy for long-tailed visual recognition called Globaland Local Mixture Consistency cumulative learning (GLMC). Our core ideas aretwofold: (1) a global and local mixture consistency loss improves therobustness of the feature extractor. Specifically, we generate two augmentedbatches by the global MixUp and local CutMix from the same batch data,respectively, and then use cosine similarity to minimize the difference. (2) Acumulative head tail soft label reweighted loss mitigates the head class biasproblem. We use empirical class frequencies to reweight the mixed label of thehead-tail class for long-tailed data and then balance the conventional loss andthe rebalanced loss with a coefficient accumulated by epochs. Our approachachieves state-of-the-art accuracy on CIFAR10-LT, CIFAR100-LT, and ImageNet-LTdatasets. Additional experiments on balanced ImageNet and CIFAR demonstratethat GLMC can significantly improve the generalization of backbones. Code ismade publicly available at https://github.com/ynu-yangpeng/GLMC.

Global and Local Mixture Consistency Cumulative Learning for Long-tailed Visual Recognitions