Zero-shot Clustering of Embeddings with Pretrained and Self-Supervised Learning Encoders
Primary Area: unsupervised, self-supervised, semi-supervised, and supervised representation learning
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Keywords: ssl, clustering
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.
Abstract: In this work, we explore whether pretrained models can provide a useful representation space for datasets they were not trained on, and whether these representations can be used to group novel unlabelled data into meaningful clusters. To this end, we conduct experiments using image representation encoders pretrained on ImageNet using either supervised or self-supervised training techniques. These encoders are deployed on image datasets that were not seen during training, and we investigate whether their embeddings can be clustered with conventional clustering algorithms. We find that it is possible to create well-defined clusters using self-supervised feature encoders, especially when using the agglomerative clustering method, and that it is possible to do so even for very fine-grained datasets such as iNaturalist. We also find indications that the Silhouette score is a good proxy of cluster quality for self-supervised feature encoders when no ground truth is available.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 8864
Loading