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
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 changed | Where | Measured | Effect |
|---|---|---|---|
| match neighbors at both local (batch) and global (overall) levels with NNM | deep clustering · CIFAR-10 | ACC: 0.816 -> 0.843 | Improved |
| match neighbors at both local (batch) and global (overall) levels with NNM | deep clustering · CIFAR-100 | ACC: 0.440 -> 0.477 | Improved |
| match neighbors at both local (batch) and global (overall) levels with NNM | deep clustering · STL-10 | ACC: 0.792 -> 0.808 | Improved |
| add the local neighbor-matching loss (ablation) | NNM · CIFAR-10 | ACC: 0.232 -> 0.843 | Improved |
| add the global neighbor-matching loss (ablation) | NNM · CIFAR-10 | ACC: 0.835 -> 0.843 | Improved |
| add the consistent loss (ablation) | NNM · CIFAR-10 | ACC: 0.826 -> 0.843 | Improved |
| add the class contrastive loss (ablation) | NNM · CIFAR-10 | ACC: 0.836 -> 0.843 | Improved |
| add the entropy loss (ablation) | NNM · CIFAR-10 | ACC: 0.204 -> 0.843 | Improved |
| increase clustering heads from 3 to 5 (ablation) | NNM · CIFAR-10 | ACC: 0.843 -> 0.845 | Improved |
| increase clustering heads from 3 to 5 (ablation) | NNM · CIFAR-100 | ACC: 0.477 -> 0.467 | Hurt |
| increase global nearest neighbors from 1 to 5 (ablation) | NNM · STL-10 | ACC: 0.808 -> 0.763 | Hurt |
| increase global nearest neighbors from 1 to 10 (ablation) | NNM · STL-10 | ACC: 0.808 -> 0.718 | Hurt |
Closest papers by meaning
- Samples Are Not Equal: A Sample Selection Approach for Deep Clustering (ICLR 2026 · similarity 0.87)
- MPCC: Matching Priors and Conditionals for Clustering (ECCV 2020 · similarity 0.87)
- DeepDPM: Deep Clustering With an Unknown Number of Clusters (CVPR 2022 · similarity 0.87)
- You Can Trust Your Clustering Model: A Parameter-free Self-Boosting Plug-in for Deep Clustering (NeurIPS 2025 · similarity 0.86)
- Revisiting Nearest Neighbor for Tabular Data: A Deep Tabular Baseline Two Decades Later (ICLR 2025 · similarity 0.84)
- Discovering Data Structures: Nearest Neighbor Search and Beyond (NeurIPS 2025 · similarity 0.84)
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.
