Research map / Label-efficient and robust learning / Recognition and long-tail
CliqueCNN: Deep Unsupervised Exemplar Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use compact cliques of mutually distant groups to construct training batches instead of training with the exemplar and its nearest neighbors (ablation) | CliqueCNN · Olympic Sports | AUC: 0.65 -> 0.79 | Helped |
| Add the exemplar's 10 nearest neighbors to training rather than using only the exemplar (ablation) | 1-sample CNN · Olympic Sports | AUC: 0.62 -> 0.65 | Helped |
| Iteratively update similarities, cliques, and batches using CNN-imputed similarities (ablation) | CliqueCNN · Olympic Sports | reliable similarities and dissimilarities per batch: 2.34 | Helped |
| Use CliqueCNN-learned similarities instead of the initial Wang et al. similarities (ablation) | Wang et al. similarities · PASCAL VOC 2007 | classification accuracy: 0.4501 -> 0.4812 | Helped |
Closest papers by meaning
- Deep Unsupervised Similarity Learning Using Partially Ordered Sets (CVPR 2017 · similarity 0.86)
- Convolutional Neural Networks With Alternately Updated Clique (CVPR 2018 · similarity 0.86)
- Unsupervised Deep Learning by Neighbourhood Discovery (ICML 2019 · similarity 0.85)
- Unsupervised Representation Learning by Predicting Image Rotations (ICLR 2018 · similarity 0.85)
- Learning Shared Representations from Unpaired Data (NeurIPS 2025 · similarity 0.83)
- unMORE: Unsupervised Multi-Object Segmentation via Center-Boundary Reasoning (ICML 2025 · similarity 0.82)
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.