Deep Co-Training for Semi-Supervised Image Recognition
This addresses the problem of learning classifiers with limited labeled data for image recognition, offering a novel deep learning adaptation of Co-Training.
The paper tackles semi-supervised image recognition by proposing Deep Co-Training, a method that trains multiple deep neural networks as different views using adversarial examples to encourage diversity, resulting in outperforming previous state-of-the-art methods by a large margin on datasets like SVHN, CIFAR-10/100, and ImageNet.
In this paper, we study the problem of semi-supervised image recognition, which is to learn classifiers using both labeled and unlabeled images. We present Deep Co-Training, a deep learning based method inspired by the Co-Training framework. The original Co-Training learns two classifiers on two views which are data from different sources that describe the same instances. To extend this concept to deep learning, Deep Co-Training trains multiple deep neural networks to be the different views and exploits adversarial examples to encourage view difference, in order to prevent the networks from collapsing into each other. As a result, the co-trained networks provide different and complementary information about the data, which is necessary for the Co-Training framework to achieve good results. We test our method on SVHN, CIFAR-10/100 and ImageNet datasets, and our method outperforms the previous state-of-the-art methods by a large margin.