AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Nearest Neighbor Matching for Deep Clustering

Zhiyuan Dang, Cheng Deng, Xu Yang, Kun Wei, Heng Huang

CVPR 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning · 100 citations (OpenAlex)

Check 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
match neighbors at both local (batch) and global (overall) levels with NNMdeep clustering · CIFAR-10ACC: 0.816 -> 0.843Improved
match neighbors at both local (batch) and global (overall) levels with NNMdeep clustering · CIFAR-100ACC: 0.440 -> 0.477Improved
match neighbors at both local (batch) and global (overall) levels with NNMdeep clustering · STL-10ACC: 0.792 -> 0.808Improved
add the local neighbor-matching loss (ablation)NNM · CIFAR-10ACC: 0.232 -> 0.843Improved
add the global neighbor-matching loss (ablation)NNM · CIFAR-10ACC: 0.835 -> 0.843Improved
add the consistent loss (ablation)NNM · CIFAR-10ACC: 0.826 -> 0.843Improved
add the class contrastive loss (ablation)NNM · CIFAR-10ACC: 0.836 -> 0.843Improved
add the entropy loss (ablation)NNM · CIFAR-10ACC: 0.204 -> 0.843Improved
increase clustering heads from 3 to 5 (ablation)NNM · CIFAR-10ACC: 0.843 -> 0.845Improved
increase clustering heads from 3 to 5 (ablation)NNM · CIFAR-100ACC: 0.477 -> 0.467Hurt
increase global nearest neighbors from 1 to 5 (ablation)NNM · STL-10ACC: 0.808 -> 0.763Hurt
increase global nearest neighbors from 1 to 10 (ablation)NNM · STL-10ACC: 0.808 -> 0.718Hurt

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