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 1–10 of 65 papers

TitleStatusHype
NeurNCD: Novel Class Discovery via Implicit Neural Representation—0
DATA: Multi-Disentanglement based Contrastive Learning for Open-World Semi-Supervised Deepfake Attribution—0
Freeze and Cluster: A Simple Baseline for Rehearsal-Free Continual Category Discovery—0
Novel Class Discovery for Open Set Raga Classification—0
AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial ScenariosCode2
OW-Rep: Open World Object Detection with Instance Representation Learning—0
Hierarchical novel class discovery for single-cell transcriptomic profilesCode0
Text3DAug -- Prompted Instance Augmentation for LiDAR PerceptionCode1
NC-NCD: Novel Class Discovery for Node ClassificationCode0
Dual-level Adaptive Self-Labeling for Novel Class Discovery in Point Cloud SegmentationCode1
Show:102550
← PrevPage 1 of 7Next →
All datasetsCIFAR10cifar100SVHN

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1AutoNovelClustering Accuracy0.75—Unverified