AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Pareto Self-Supervised Training for Few-Shot Learning

Zhengyu Chen, Jixie Ge, Heshen Zhan, Siteng Huang, Donglin Wang

CVPR 2021 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Data-efficient and robust self-supervision · 147 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 PSST with rotation prediction instead of the prior auxiliary-training setupMetaOptNet · MiniImageNet 1-shotaccuracy: 62.93 ± 0.63 -> 64.05± 0.49Improved
Use PSST with rotation prediction instead of the prior auxiliary-training setupMetaOptNet · MiniImageNet 5-shotaccuracy: 79.63 ± 0.47 -> 80.24 ± 0.45Improved
Use PSST with location prediction instead of BF3S auxiliary trainingCC · MiniImageNet 1-shotaccuracy: 60.71 ± 0.46 -> 63.26 ± 0.44Improved
Use PSST with location prediction instead of BF3S auxiliary trainingCC · MiniImageNet 5-shotaccuracy: 77.64 ± 0.34 -> 80.05 ± 0.34Improved
Use PSST with rotation prediction instead of BF3S auxiliary trainingCC · MiniImageNet 1-shotaccuracy: 62.93 ± 0.45 -> 64.16 ± 0.44Improved
Use PSST with rotation prediction instead of BF3S auxiliary trainingCC · MiniImageNet 5-shotaccuracy: 79.87 ± 0.33 -> 80.64 ± 0.32Improved
Use PSST with rotation prediction instead of SSFSL auxiliary trainingProtoNet · MiniImageNet 1-shotaccuracy: 57.92 ± 0.58 -> 59.28 ± 0.46Improved
Use PSST with jigsaw prediction instead of SSFSL auxiliary trainingProtoNet · MiniImageNet 5-shotaccuracy: 76.20 ± 0.53 -> 77.27 ± 0.43Improved
Add location prediction without PSST (ablation)CC · MiniImageNet 1-shotaccuracy: 61.09 ± 0.44 -> 60.71 ± 0.46Hurt
Add location prediction without PSST (ablation)CC · MiniImageNet 5-shotaccuracy: 78.43 ± 0.33 -> 77.64 ± 0.34Hurt
Use PSST instead of grid-search weights to obtain multiple trade-off solutionsBF3S · Training, 10 solutionsiterations: 249 -> 60Improved
Restrict Pareto exploration to the preferred regionPareto exploration on MultiMNIST · MultiMNISTiterations: 130 -> 78Improved

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