AI Paper Atlas

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

HCSC: Hierarchical Contrastive Selective Coding

Yuanfan Guo, Minghao Xu, Jiawen Li, Bingbing Ni, Xuanyu Zhu, Zhenbang Sun, Yi Xu

CVPR 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning · 85 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
Use hierarchical prototypes and selective instance-wise and prototypical contrastive learningHCSC · ImageNet linear classificationTop1-Acc: 68.6 -> 69.2Improved
Add hierarchical prototypes to instance-wise and prototypical contrastive losses (ablation)HCSC · ImageNet linear classificationTop1-Acc: 67.6 -> 68.1Improved
Add instance-wise pair selection (ablation)HCSC · ImageNet linear classificationTop1-Acc: 68.1 -> 68.9Improved
Add prototypical pair selection alone (ablation)HCSC · ImageNet linear classificationTop1-Acc: 68.1 -> 68.2Improved
Combine instance-wise and prototypical pair selection (ablation)HCSC · ImageNet linear classificationTop1-Acc: 68.9 -> 69.2Improved
Add instance-wise contrastive loss (ablation)HCSC · ImageNet linear classificationTop1-Acc: 65.7 -> 69.2Improved
Add prototypical contrastive loss (ablation)HCSC · ImageNet linear classificationTop1-Acc: 68.2 -> 69.2Improved
Increase prototype hierarchy count from one to two (ablation)HCSC · ImageNet linear classificationTop1-Acc: 68.0 -> 69.0Improved
Increase prototype hierarchy count from three to four (ablation)HCSC · ImageNet linear classificationTop1-Acc: 69.2 -> 69.2No clear effect
Use insufficient prototype counts, 1000-500-200, rather than 3000-2000-1000 (ablation)HCSC · ImageNet linear classificationTop1-Acc: 69.2 -> 68.7Hurt
Use initial learning rate 30.0 in downstream linear evaluationHCSC · ImageNet linear classificationTop1-Acc: 67.5 -> 67.3Hurt
Pretrain with HCSC rather than AdCoImageNet-pretrained encoder · CUB zero-shotKNN-Top1-Acc: 22.9 -> 26.9Improved

Later papers that compared with this method

1 later accepted paper reports HCSC as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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