AI Paper Atlas

Research map / Label-efficient and robust learning / Domain adaptation and generalization

Rethinking Normalization Layers for Domain Generalizable Person Re-identification

Ren Nie, Jin Ding, Xue Zhou, Xi Li

ECCV 2024 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Generalizable person re-identification · 13 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 Domain Frozen (freeze normalization affine parameters) (ablation)ResNet50 baseline · Protocol-2 averagemAP: 40.5 -> 42.3Improved
add Remix Normalization (IN+BN statistics averaged) (ablation)ResNet50 baseline · Protocol-2 averagemAP: 40.5 -> 43.9Improved
add Emulation Normalization on top of RN+DF (ablation)ReNorm · Protocol-2 averagemAP: 46.0 -> 47.3Improved
add Domain Frozen on top of RN+EN (ablation)ReNorm · Protocol-2 averagemAP: 45.3 -> 47.3Improved
plug ReNorm into all four ResNet50 stages instead of stage 4 only (ablation)ReNorm · Protocol-2 target CUHK03mAP: 39.4 -> 43.6Improved
average Network-R and Network-E outputs rather than using one network (ablation)ReNorm inference · Protocol-2 averagemAP: 46.3 -> 47.3Improved
average instead of concatenate the R and E features (ablation)ReNorm inference · Protocol-2 averagemAP: 47.0 -> 47.3Improved
ReNorm vs best prior method on Protocol-1ReNorm · Protocol-1 averagemAP: 76.2 -> 78.2Improved
ReNorm on iLIDs target versus prior methodsReNorm · Protocol-1 iLIDsmAP: 87.0 vs 89.9 (SNR)Hurt

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