论文标题
R2-D2：重复拒绝的深度解密，用于半监督的深度学习
R2-D2: Repetitive Reprediction Deep Decipher for Semi-Supervised Deep Learning
论文作者
论文摘要
最新的半监督深度学习（深SSL）方法使用了类似的范式：使用网络预测来更新伪标签，并使用伪标记来迭代更新网络参数。但是，他们缺乏理论支持，无法解释为什么预测是深度学习范式中伪标记的良好候选者。在本文中，我们提出了一个原则上的端到端框架，称为SSL的Deep Degipher（D2）。在D2框架中，我们证明了伪标记与指数链接函数与网络预测有关，该链接函数为使用预测作为伪标记提供了理论支持。此外，我们证明，通过网络预测更新伪标签将使它们不确定。为了减轻这个问题，我们提出了一种称为重复拒绝（R2）的培训策略。最后，在大型成像网数据集上测试了提出的R2-D2方法，并以5个百分点的优于最先进的方法。
Most recent semi-supervised deep learning (deep SSL) methods used a similar paradigm: use network predictions to update pseudo-labels and use pseudo-labels to update network parameters iteratively. However, they lack theoretical support and cannot explain why predictions are good candidates for pseudo-labels in the deep learning paradigm. In this paper, we propose a principled end-to-end framework named deep decipher (D2) for SSL. Within the D2 framework, we prove that pseudo-labels are related to network predictions by an exponential link function, which gives a theoretical support for using predictions as pseudo-labels. Furthermore, we demonstrate that updating pseudo-labels by network predictions will make them uncertain. To mitigate this problem, we propose a training strategy called repetitive reprediction (R2). Finally, the proposed R2-D2 method is tested on the large-scale ImageNet dataset and outperforms state-of-the-art methods by 5 percentage points.
