Abstract
Self-training is a powerful approach to deep learning. The key process is to find a pseudo-label for modeling. However, previous self-training algorithms suffer from the over-confidence issue brought by the hard labels, even some confidence-related regularizers cannot comprehensively catch the uncertainty. Therefore, we propose a new self-training framework to combine uncertainty information of both model and dataset. Specifically, we propose to use Expectation-Maximization (EM) to smooth the labels and comprehensively estimate the uncertainty information. We further design a basis extraction network to estimate the initial basis from the dataset. The obtained basis with uncertainty can be filtered based on uncertainty information. It can then be transformed into the real hard label to iteratively update the model and basis in the retraining process. Experiments on image classification and semantic segmentation show the advantages of our methods among confidence-aware self-training algorithms with 1-3 percentage improvement on different datasets.
Abstract (translated)
自我训练是一种强大的深度学习方法。关键的过程是找到一个伪标签来建模。然而,之前的自我训练算法由于硬标签带来的过度自信问题,即使是一些与信心相关的正则化器也无法全面捕捉不确定性。因此,我们提出了一个新的自我训练框架来结合模型和数据的不确定性信息。具体来说,我们提出使用Expectation-Maximization(EM)来平滑标签,并全面估计不确定性信息。我们还设计了一个基提取网络,从数据集中估计初始基。具有不确定性的基可以基于不确定性信息进行筛选。然后可以将其转换为真实硬标签,在重新训练过程中逐步更新模型和基。在图像分类和语义分割上的实验表明,我们的方法在关注信心的自训练算法中具有1-3个百分比的改进。
URL
https://arxiv.org/abs/2405.01175
