Deep Clustering for Unsupervised Learning of Visual Features
This addresses the problem of unsupervised learning for computer vision researchers, offering a novel method that is not incremental but advances the field.
The paper tackles unsupervised visual feature learning by introducing DeepCluster, which jointly trains neural network parameters and cluster assignments using k-means, and achieves state-of-the-art results on benchmarks like ImageNet and YFCC100M with significant performance margins.
Clustering is a class of unsupervised learning methods that has been extensively applied and studied in computer vision. Little work has been done to adapt it to the end-to-end training of visual features on large scale datasets. In this work, we present DeepCluster, a clustering method that jointly learns the parameters of a neural network and the cluster assignments of the resulting features. DeepCluster iteratively groups the features with a standard clustering algorithm, k-means, and uses the subsequent assignments as supervision to update the weights of the network. We apply DeepCluster to the unsupervised training of convolutional neural networks on large datasets like ImageNet and YFCC100M. The resulting model outperforms the current state of the art by a significant margin on all the standard benchmarks.