AI Paper Atlas

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

Weakly Supervised Representation Learning With Coarse Labels

Yuanhong Xu, Qi Qian, Hao Li, Rong Jin, Juhua Hu

ICCV 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive representation theory · 11 citations (OpenAlex)

Read the paperCheck 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
add instance classification to coarse-class training (CoIns vs Cos)ResNet-18 · CIFAR-100 (100 target classes)Recall@1: 37.1 -> 57.0Improved
add coarse-class classification to instance classification (CoIns vs Ins)ResNet-18 · CIFAR-100 (100 target classes)Recall@1: 13.6 -> 57.0Improved
add instance classification to coarse-class training (CoIns vs Cos)ResNet-18 · SOP (13,160 target classes)Recall@1: 21.8 -> 35.8Improved
add instance classification to coarse labels (CoIns vs Cos)ResNet-50 MoCo-v2 · ImageNet (1000 target classes)Recall@1: 21.1 -> 51.4Improved
add coarse labels to MoCo-v2 instance classificationResNet-50 · ImageNet (1000 target classes, linear probe)Top-1 accuracy: 67.5 -> 70.4Improved
add instance classification for coarse-class accuracyResNet-18 · CIFAR-100 (20 coarse classes)Top-1 accuracy: 85.6 -> 86.3Improved
restrict instance classification within each coarse class (CoIns_imp vs CoIns)ResNet-18 · CIFAR-100 (100 target classes)Recall@1: 57.0 -> 56.6No clear effect
add cosine softmaxCoIns · CIFAR-100 (100 target classes)Recall@1: 57.0 -> 60.8Improved
add MLP head on top of cosine softmaxCoIns* · CIFAR-100 (100 target classes)Recall@1: 60.8 -> 60.5No clear effect
add MLP head on top of cosine softmaxCoIns* · SOP (13,160 target classes)Recall@1: 38.1 -> 42.7Improved
add strong augmentationCoIns · CIFAR-100RecallHurt
add instance proxy loss at epoch 100 (ablation)CoIns** · CIFAR-100 (100 target classes)Recall@1: 60.5 -> 62.0Improved

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