[论文解读] To Balance or Not to Balance: An Embarrassingly Simple Approach for Learning with Long-Tailed Distributions.

Junjie Zhang, Lingqiao Liu+2 more|arXiv (Cornell University)|Dec 10, 2019

Anomaly Detection Techniques and Applications被引用 2

一句话总结

该论文提出了一种简单但有效的方法，通过将网络解耦为特征提取器和分类器，对两者均采用类别平衡采样，并在随机采样下为特征提取器引入辅助自监督任务，从而在长尾视觉数据集上训练深度神经网络。该方法不依赖复杂组件，通过缓解过拟合并提升尾部类别泛化能力，实现了最先进性能。

ABSTRACT

Real-world visual data often exhibits a long-tailed distribution, where some ''head'' classes have a large number of samples, yet only a few samples are available for ''tail'' classes. Such imbalanced distribution causes a great challenge for learning a deep neural network, which can be boiled down into a dilemma: on the one hand, we prefer to increase the exposure of tail class samples to avoid the excessive dominance of head classes in the classifier training. On the other hand, oversampling tail classes makes the network prone to over-fitting, since head class samples are often consequently under-represented. To resolve this dilemma, in this paper, we propose a simple-yet-effective auxiliary learning approach. The key idea is to split a network into a classifier part and a feature extractor part, and then employ different training strategies for each part. Specifically, to promote the awareness of tail-classes, a class-balanced sampling scheme is utilised for training both the classifier and the feature extractor. For the feature extractor, we also introduce an auxiliary training task, which is to train a classifier under the regular random sampling scheme. In this way, the feature extractor is jointly trained from both sampling strategies and thus can take advantage of all training data and avoid the over-fitting issue. Apart from this basic auxiliary task, we further explore the benefit of using self-supervised learning as the auxiliary task. Without using any bells and whistles, our model achieves superior performance over the state-of-the-art solutions.

研究动机与目标

• 解决真实世界视觉数据集中头部类别占主导、尾部类别代表性不足的长尾类别分布挑战。
• 解决增加尾部类别曝光与因过度采样导致过拟合之间的训练困境。
• 在不依赖复杂数据增强或架构修改的前提下，提升模型在罕见（尾部）类别上的泛化能力。
• 探索通过为特征提取器引入辅助学习任务，结合类别平衡采样与随机采样的有效性。

提出的方法

• 将网络拆分为特征提取器和分类器，使每个组件可独立应用不同的训练策略。
• 对分类器和特征提取器均使用类别平衡采样，以增强对代表性不足的尾部类别的感知能力。
• 为特征提取器引入一个辅助训练任务，采用常规随机采样，使其能够从全部数据中学习并避免对稀有类别的过拟合。
• 通过两种采样方案联合优化特征提取器，结合平衡采样与全量数据暴露的优势。
• 探索自监督学习作为辅助任务，以进一步提升特征质量，且无需额外标签。
• 该方法不依赖复杂组件或附加功能，仅依靠策略性采样和辅助监督。

实验结果

研究问题

• RQ1 是否可以通过使用不同采样方案的简单双训练策略，提升长尾视觉识别任务的性能？
• RQ2 通过辅助任务将类别平衡采样与随机采样结合，能否缓解尾部类别学习中的过拟合问题？
• RQ3 在类别不平衡设置下，自监督学习作为辅助任务在多大程度上能增强特征表示？
• RQ4 与最先进方法相比，该方法在不使用复杂数据或模型工程的情况下表现如何？

主要发现

• 所提方法在长尾视觉基准上实现了最先进性能，且未使用复杂数据增强或模型修改。
• 对分类器和特征提取器均采用类别平衡采样，显著提升了尾部类别识别准确率。
• 在随机采样下引入的辅助训练任务有助于特征提取器更好地泛化，避免对稀有类别的过拟合。
• 作为辅助任务的自监督学习进一步提升了性能，证明了在类别不平衡设置下无监督预训练的有效性。
• 即使不依赖损失重加权或先进数据采样技术等附加组件，该方法仍优于现有方法。