AI Paper Atlas

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

A Unified Approach to Kinship Verification

Eran Dahan, Yosi Keller

TPAMI 2020 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 23 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 the 1x1-convolution embedding fusion subnetwork instead of concatenation plus FC (ablation)unified kin verification network · RFIW unrestrictedaverage accuracy (%): 55.65 -> 79.55Improved
add the fusion subnetwork on top of raw face embeddings (ablation)unified kin verification network · RFIW unrestrictedaverage accuracy (%): 73.65 -> 79.55Improved
add adaptive family-balanced sampling instead of random sampling (ablation)unified kin verification network · RFIW unrestrictedaverage accuracy (%): 75.00 -> 79.55Improved
add multi-task training across all kinship classes instead of one CNN per class (ablation)unified kin verification network · RFIW unrestrictedaverage accuracy (%): 78.42 -> 79.55Improved
add asymmetric embedding weighting vs no normalization (ablation)unified kin verification network · RFIW unrestrictedaverage accuracy (%): 78.90 -> 79.55Improved
use asymmetric weights instead of symmetric weights (ablation)unified kin verification network · RFIW unrestrictedaverage accuracy (%): 78.98 -> 79.55Improved
vary the BCE loss weight alpha (ablation)unified kin verification network · RFIWaccuracyNo clear effect
switch backbone from VGG-Face to fine-tuned SphereFaceproposed fusion scheme · RFIW FG2018average accuracy (%): 71.1 -> 79.6Improved
add the proposed fusion and classification network on top of SphereFaceSphereFace · RFIW FG2018average accuracy (%): 71.2 -> 79.6Improved
use proposed scheme versus best distance-learning baseline on VGG-FaceVGG-Face · RFIW FG2018average accuracy (%): 68.5 -> 71.1Improved
use proposed scheme versus FG2020 top entry on grandparent-free classesFG2020 challenge entries · FG2020 (unrestricted)average accuracy: 0.77 -> 0.78Mixed

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