监督|长尾问题太严重？半监督和自监督就可以有效缓解！( 五 ) 不平衡|数据集|样本|缓解

拿一位给我们很高分数的reviewer的原话，“The results could be of interest to even broader area of different applications”，即不只是局限于文中做的几个academic datasets，而对于现实中许多常见的imbalance或long-tail的任务，都是能即插即用，或是对如何有效收集无标签数据提供一些insight的。
当然，宣传归宣传，我们的工作还是存在其局限性。虽然我们考虑到了无标签数据的不平衡性，但是对于半监督（或是自监督）的算法本身，并没有整合不平衡学习的策略，而是直接使用了vanilla的算法。其次，如我们标题所带词语“improving”所示，我们能提升现有的最优算法，但长尾问题本身仍未完全解决，甚至还有很大的提升空间。
参考文献：
[1]Yin Cui, Menglin Jia, Tsung-Yi Lin, Yang Song, and Serge Belongie. Class-balanced loss based on effective number of samples. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 9268–9277, 2019.
[2]Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Dollár. Focal loss for dense object detection. In ICCV, pages 2980–2988, 2017.[3]Samira Pouyanfar, et al. Dynamic sampling in convolutional neural networks for imbalanced data classification.[4]Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss. NeurIPS, 2019.[5]BBN: Bilateral-Branch Network with Cumulative Learning for Long-Tailed Visual Recognition. CVPR, 2020.[6]Decoupling representation and classifier for long-tailed recognition. ICLR, 2020.[7]iNatrualist 2018 competition dataset. https://github.com/visipedia/inat_comp/tree/master/2018[8]He, H. and Garcia, E. A. Learning from imbalanced data. TKDE, 2008.[9]Chawla, N. V., et al. SMOTE: synthetic minority oversampling technique. JAIR, 2002.[10]mixup: Beyond Empirical Risk Minimization. ICLR 2018.[11]H. Chou et al. Remix: Rebalanced Mixup. 2020.[12]Deep Imbalanced Learning for Face Recognition and Attribute Prediction. TPAMI, 2019.[13]Large-scale long-tailed recognition in an open world. CVPR, 2019.[14]Feature transfer learning for face recognition with under-represented data. CVPR, 2019.[15]Range Loss for Deep Face Recognition with Long-Tail. CVPR, 2017.[16]Learning Deep Representation for Imbalanced Classification. CVPR, 2016.[17]Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting. NeurIPS, 2019.[18]Rethinking Class-Balanced Methods for Long-Tailed Recognition from a Domain Adaptation Perspective. CVPR, 2020.[19]Spyros Gidaris, Praveer Singh, and Nikos Komodakis. Unsupervised representation learning by predicting image rotations. arXiv preprint arXiv:1803.07728, 2018.[20]Kaiming He, Haoqi Fan, Yuxin Wu, Saining Xie, and Ross Girshick. Momentum contrast for unsupervised visual representation learning. arXiv preprint arXiv:1911.05722, 2019.