Interpolation Consistency Training for Semi-supervised Learning
Interpolation Consistency Training for Semi-supervised Learning
Vikas Verma, Alex Lamb, Juho Kannala, Yoshua Bengio, David Lopez-Paz
Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence
Main track. Pages 3635-3641.
https://doi.org/10.24963/ijcai.2019/504
We introduce Interpolation Consistency Training (ICT), a simple and computation efficient algorithm for training Deep Neural Networks in the semi-supervised learning paradigm. ICT encourages the prediction at an interpolation of unlabeled points to be consistent with the interpolation of the predictions at those points. In classification problems, ICT moves the decision boundary to low-density regions of the data distribution. Our experiments show that ICT achieves state-of-the-art performance when applied to standard neural network architectures on the CIFAR-10 and SVHN benchmark dataset.
Keywords:
Machine Learning: Semi-Supervised Learning
Machine Learning: Deep Learning