SOTAVerified

Novel Class Discovery

The goal of Novel Class Discovery (NCD) is to identify new classes in unlabeled data, by exploiting prior knowledge from known classes. In this specific setup, the data is split in two sets. The first is a labeled set containing known classes and the second is an unlabeled set containing unknown classes that must be discovered.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 31–40 of 65 papers

TitleStatusHype
Adaptive Discovering and Merging for Incremental Novel Class Discovery—0
Beyond the Known: Novel Class Discovery for Open-world Graph Learning—0
Boosting Novel Category Discovery Over Domains with Soft Contrastive Learning and All-in-One Classifier—0
Continual Novel Class Discovery via Feature Enhancement and Adaptation—0
DATA: Multi-Disentanglement based Contrastive Learning for Open-World Semi-Supervised Deepfake Attribution—0
Debiased Novel Category Discovering and Localization—0
Exclusive Style Removal for Cross Domain Novel Class Discovery—0
Federated Continual Novel Class Learning—0
Freeze and Cluster: A Simple Baseline for Rehearsal-Free Continual Category Discovery—0
Mutual Information-guided Knowledge Transfer for Novel Class Discovery—0
Show:102550
← PrevPage 4 of 7Next →
All datasetsCIFAR10cifar100SVHN

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1AutoNovelClustering Accuracy0.92—Unverified
#ModelMetricClaimedVerifiedStatus
1AutoNovelClustering Accuracy0.75—Unverified
#ModelMetricClaimedVerifiedStatus
1AutoNovelClustering Accuracy0.95—Unverified