Self-labelling via simultaneous clustering and representation learning
This addresses the challenge of training competitive image representations without manual labels, which is crucial for reducing annotation costs in computer vision, though it builds on existing clustering and representation learning approaches.
The paper tackles the problem of unsupervised learning for deep neural networks by proposing a method that simultaneously clusters and learns representations to self-label visual data, achieving state-of-the-art performance on datasets like SVHN, CIFAR-10, CIFAR-100, and ImageNet, with the first self-supervised AlexNet outperforming the supervised Pascal VOC detection baseline.
Combining clustering and representation learning is one of the most promising approaches for unsupervised learning of deep neural networks. However, doing so naively leads to ill posed learning problems with degenerate solutions. In this paper, we propose a novel and principled learning formulation that addresses these issues. The method is obtained by maximizing the information between labels and input data indices. We show that this criterion extends standard crossentropy minimization to an optimal transport problem, which we solve efficiently for millions of input images and thousands of labels using a fast variant of the Sinkhorn-Knopp algorithm. The resulting method is able to self-label visual data so as to train highly competitive image representations without manual labels. Our method achieves state of the art representation learning performance for AlexNet and ResNet-50 on SVHN, CIFAR-10, CIFAR-100 and ImageNet and yields the first self-supervised AlexNet that outperforms the supervised Pascal VOC detection baseline. Code and models are available.