AI Paper Atlas

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

Decoupled Contrastive Learning

Chun-Hsiao Yeh, Cheng-Yao Hong, Yen-Chi Hsu, Tyng-Luh Liu, Yubei Chen, Yann LeCun

ECCV 2022 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Contrastive versus non-contrastive theory · 235 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
remove positive term from InfoNCE denominator (DCL) (ablation)SimCLR · ImageNet-1K (bs256, 200 ep)linear top-1: 61.8 -> 65.9Improved
apply DCL at small batch size 32 (ablation)SimCLR · ImageNet-1K (bs32)linear top-1: 56.8 -> 61.5Improved
apply DCL at batch size 512 (ablation)SimCLR · ImageNet-1K (bs512)linear top-1: 64.0 -> 65.8Improved
apply DCL at batch size 1024 (ablation)SimCLR · ImageNet-1K (bs1024)linear top-1: 65.1 -> 66.1Improved
add DCL to MoCo-v2 with small queue 64 (ablation)MoCo-v2 · ImageNet-1Klinear top-1: 63.9 -> 65.8Improved
add DCL to MoCo-v2 with queue 65536 (ablation)MoCo-v2 · ImageNet-1Klinear top-1: 67.5 -> 67.7No clear effect
add von Mises-Fisher positive weighting (DCLW) (ablation)DCL · ImageNet-1K (bs256, 200 ep)linear top-1: 65.9 -> 66.9Improved
tune temperature and learning rate (ablation)DCL · ImageNet-1Klinear top-1: 65.9 -> 67.8Improved
add BYOL asymmetric augmentation (ablation)DCL · ImageNet-1Klinear top-1: 67.8 -> 68.2Improved
add DCL when training only 100 epochs (ablation)SimCLR · ImageNet-1K (100 ep)linear top-1: 57.5 -> 64.6Improved
combine DCL with NNCLR at batch size 512 (ablation)NNCLR · ImageNet-1Klinear top-1: 71.7 -> 72.3Improved
use DCL over InfoNCE for temperature sensitivity (ablation)SimCLR · CIFAR10 temperature sweepstd of kNN accuracy: 1.44 -> 0.98Improved

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