AI Paper Atlas

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

Pairwise Relational Networks for Face Recognition

Bong-Nam Kang, Yonghyun Kim, Daijin Kim

ECCV 2018 · Label-efficient and robust learning › Recognition and long-tail · approach: Margin-based face recognition losses · 42 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 face identity state feature to the PRN (ablation)PRN · VGGFace2 validation setclassification accuracy: 94.2 -> 96.7Improved
add PRN (no identity state) to the global CNN feature (ablation)model A · LFWaccuracy (%): 99.6 -> 99.65Improved
add identity state feature to the PRN branch (ablation)model B · LFWaccuracy (%): 99.65 -> 99.76Improved
add PRN+ to the global CNN feature (ablation)model A · LFWaccuracy (%): 99.6 -> 99.76Improved
add PRN (no identity state) to the global CNN feature (ablation)model A · YTFaccuracy (%): 95.1 -> 95.7Improved
add identity state feature to the PRN branch (ablation)model B · YTFaccuracy (%): 95.7 -> 96.3Improved
add PRN+ to the global CNN feature (ablation)model A · YTFaccuracy (%): 95.1 -> 96.3Improved
add identity state feature to the PRN branch (ablation)model B · IJB-A 1:1 verification FAR=0.001TAR: 0.901 -> 0.919Improved
add PRN+ to the global CNN feature (ablation)model A · IJB-A 1:1 verification FAR=0.001TAR: 0.895 -> 0.919Improved
add PRN (no identity state) to the global CNN feature (ablation)model A · IJB-A 1:N identificationRank-1 TPIR: 0.975 -> 0.976No clear effect
add identity state feature to the PRN branch (ablation)model B · IJB-B 1:1 verification FAR=0.001TAR: 0.910 -> 0.923Improved
add PRN+ to the global CNN feature (ablation)model A · IJB-B 1:1 verification FAR=0.001TAR: 0.892 -> 0.923Improved

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