Zhun Zhong
Enrico Fini
Subhankar Roy
Zhiming Luo
Elisa Ricci
Nicu Sebe
2021-06-22
<p>In this paper, we address Novel Class Discovery (NCD), the task of unveiling new classes in a set of unlabeled samples given a labeled dataset with known classes. We exploit the peculiarities of NCD to build a new framework, named Neighborhood Contrastive Learning (NCL), to learn discriminative representations that are important to clustering performance. Our contribution is twofold. First, we find that a feature extractor trained on the labeled set generates representations in which a generic query sample and<br>
its neighbors are likely to share the same class. We exploit this observation to retrieve and aggregate pseudo-positive pairs with contrastive learning, thus encouraging the model to learn more discriminative representations. Second, we notice that most of the instances are easily discriminated by the network, contributing less to the contrastive loss. To overcome this issue, we propose to generate hard negatives by mixing labeled and unlabeled samples in the feature space. We experimentally demonstrate that these two ingredients significantly contribute to clustering performance and lead our model to outperform state-of-the-art methods by a large margin (e.g., clustering accuracy +13% on CIFAR-100 and +8% on ImageNet).</p>
IEEE/CVF International Conference on Computer Vision and Pattern Recognition (CVPR'21)
https://doi.org/10.5281/zenodo.5014108
oai:zenodo.org:5014108
Zenodo
https://zenodo.org/communities/ai4media
https://doi.org/10.5281/zenodo.5014107
info:eu-repo/semantics/openAccess
Creative Commons Attribution 4.0 International
https://creativecommons.org/licenses/by/4.0/legalcode
CVPR'21, IEEE/CVF International Conference on Computer Vision and Pattern Recognition
Neighborhood Contrastive Learning for Novel Class Discovery
info:eu-repo/semantics/conferencePaper