AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

Graph Contrastive Clustering

Huasong Zhong, Jianlong Wu, Chong Chen, Jianqiang Huang, Minghua Deng, Liqiang Nie, Zhouchen Lin, Xian-Sheng Hua

ICCV 2021 · Graph learning › Graph self-supervised learning · approach: Contrastive learning with negatives and views · 167 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add representation graph contrastive loss (ablation)GCC without graph contrastive · CIFAR-10ACC: 0.752 -> 0.809Improved
add assignment graph contrastive loss (ablation)GCC without graph contrastive · CIFAR-10ACC: 0.752 -> 0.825Improved
add both RGC and AGC (ablation)GCC without graph contrastive · CIFAR-10ACC: 0.752 -> 0.856Improved
add both RGC and AGC (ablation)GCC without graph contrastive · CIFAR-100ACC: 0.438 -> 0.472Improved
add both RGC and AGC (ablation)GCC without graph contrastive · ImageNet-10ACC: 0.878 -> 0.901Improved
add cluster regularization loss (ablation)GCC w/o CR · CIFAR-10ACC: 0.680 -> 0.856Improved
add cluster regularization loss (ablation)GCC w/o CR · CIFAR-100ACC: 0.348 -> 0.472Improved
add cluster regularization loss (ablation)GCC w/o CR · ImageNet-10ACC: 0.828 -> 0.901Improved
apply SCAN self-labeling fine-tuning to GCCGCC · CIFAR-10ACC: 0.856 -> 0.901Improved
use GCC instead of SCAN, with self-labelingSCAN* · CIFAR-100ACC: 0.507 -> 0.523Improved
use graph contrastive features instead of simCLR (spectral clustering)simCLR + SC · CIFAR-10ACC: 0.660 -> 0.746Improved
use graph contrastive features instead of simCLR (K-means)simCLR + K-means · CIFAR-100ACC: 0.380 -> 0.420Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy