Consensus Clustering With Unsupervised Representation Learning
This work addresses a specific limitation in unsupervised representation learning for clustering, but it is incremental as it builds on existing BYOL and clustering techniques.
The authors tackled the problem that features learned by the BYOL self-supervised method may not be optimal for clustering, and they proposed a consensus clustering loss to improve clustering performance, achieving better results than similar methods on some computer vision datasets.
Recent advances in deep clustering and unsupervised representation learning are based on the idea that different views of an input image (generated through data augmentation techniques) must either be closer in the representation space, or have a similar cluster assignment. Bootstrap Your Own Latent (BYOL) is one such representation learning algorithm that has achieved state-of-the-art results in self-supervised image classification on ImageNet under the linear evaluation protocol. However, the utility of the learnt features of BYOL to perform clustering is not explored. In this work, we study the clustering ability of BYOL and observe that features learnt using BYOL may not be optimal for clustering. We propose a novel consensus clustering based loss function, and train BYOL with the proposed loss in an end-to-end way that improves the clustering ability and outperforms similar clustering based methods on some popular computer vision datasets.