Spacing Loss for Discovering Novel Categories
Abstract
Novel Class Discovery (NCD) is a learning paradigm, where a machine learning model is tasked to semantically group instances from unlabeled data, by utilizing labeled instances from a disjoint set of classes. In this work, we first characterize existing NCD approaches into singlestage and two-stage methods based on whether they require access to labeled and unlabeled data together while discovering new classes. Next, we devise a simple yet powerful loss function that enforces separability in the latent space using cues from multi-dimensional scaling, which we refer to as Spacing Loss. Our proposed formulation can either operate as a standalone method or can be plugged into existing methods to enhance them. We validate the efficacy of Spacing Loss with thorough experimental evaluation across multiple settings on CIFAR-10 and CIFAR-100 datasets.
Cite
Text
Joseph et al. "Spacing Loss for Discovering Novel Categories." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022. doi:10.1109/CVPRW56347.2022.00420Markdown
[Joseph et al. "Spacing Loss for Discovering Novel Categories." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022.](https://mlanthology.org/cvprw/2022/joseph2022cvprw-spacing/) doi:10.1109/CVPRW56347.2022.00420BibTeX
@inproceedings{joseph2022cvprw-spacing,
title = {{Spacing Loss for Discovering Novel Categories}},
author = {Joseph, K. J. and Paul, Sujoy and Aggarwal, Gaurav and Biswas, Soma and Rai, Piyush and Han, Kai and Balasubramanian, Vineeth N.},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
year = {2022},
pages = {3760-3765},
doi = {10.1109/CVPRW56347.2022.00420},
url = {https://mlanthology.org/cvprw/2022/joseph2022cvprw-spacing/}
}