AI Paper Atlas

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

ANFI: Rethinking Neighbor Feature Interaction in Person Re-ID

Xulin Li, Yan Lu, Bin Liu, Jiaze Li, Qinhong Yang, Tao Gong, Qi Chu, Nenghai Yu

ECCV 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Supervised person re-identification · 0 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 affinity relations over single-image features (ablation)ANFI baseline · Market1501 StandardmAP: 90.3 -> 94.2Improved
add affinity relations over single-image features (ablation)ANFI baseline · Market1501 1-shotmAP: 93.5 -> 92.5Hurt
add affinity relations over single-image features (ablation)ANFI baseline · Market1501 Hard QuerymAP: 65.5 -> 63.1Hurt
add discrepancy relations to affinity relations (ablation)ANFI · Market1501 1-shotmAP: 92.5 -> 93.5Improved
add discrepancy relations to affinity relations (ablation)ANFI · Market1501 Hard QuerymAP: 63.1 -> 65.3Improved
add discrepancy relations to affinity relations (ablation)ANFI · Market1501 StandardmAP: 94.2 -> 94.8Improved
add noise simulation in training (ablation)ANFI · Market1501 StandardmAP: 94.8 -> 95.1Improved
add relationship regularization Lrr (ablation)ANFI · Market1501 StandardmAP: 95.1 -> 95.7Improved
add relationship regularization Lrr (ablation)ANFI · Market1501 1-shotmAP: 93.8 -> 94.4Improved
add relationship regularization Lrr (ablation)ANFI · Market1501 Hard QuerymAP: 65.9 -> 66.1Improved
replace learned adaptive weight w_i with test-time oracle nnr (ablation)ANFI · Market1501 Standard and Hard QuerymAP: -0.1 Standard, -0.3 Hard QueryHurt
use ANFI instead of affinity-based re-ranking and neighbor methodsANFI · Market1501 Standarddelta mAP: +5.4 vs +4.9 CA-Jacc., +4.8 K-recip.Improved

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