AI Paper Atlas

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

Exploring Long Tail Visual Relationship Recognition With Large Vocabulary

Sherif Abdelkarim, Aniket Agarwal, Panos Achlioptas, Jun Chen, Jiaji Huang, Boyang Li, Kenneth Church, Mohamed Elhoseiny

ICCV 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Decoupled and calibrated long-tail training · 24 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 hub100k to the baseline classifierbaseline · GQA tripletoverall triplet accuracy: 14.7 -> 15.3Improved
add hub100k to the baseline classifierbaseline · VG8K tripletoverall triplet accuracy: 14.2 -> 15.4Improved
add hub100k to the baseline classifierbaseline · GQA triplet, thresholds (20, 100) with synsetsfew-shot triplet score: 12.6 -> 15.5Improved
add hub100k to the baseline classifierbaseline · GQA triplet, thresholds (20, 100) with synsetsmany-shot triplet score: 63.2 -> 59.8Hurt
increase hub size from hub1k to hub100k (ablation)hub variants · GQA triplet, thresholds (20, 100) with synsetsall-class triplet score: 13.1 -> 16.0Improved
increase hub size from hub100k to hub500k (ablation)hub variants · GQA triplet, thresholds (20, 100) with synsetsall-class triplet score: 16.0 -> 16.0No clear effect
add hub20k to weighted cross-entropyw_ce · GQA tripletoverall triplet accuracy: 5.3 -> 4.7Hurt
add hub100k to dcpldcpl · GQA tripletoverall triplet accuracy: 11.1 -> 9.5Hurt
add hub100k to fcfc · GQA tripletoverall triplet accuracy: 13.4 -> 14.0Improved
add hub20k to focal loss (gamma 5)focal_5 · GQA tripletoverall triplet accuracy: 14.7 -> 15.3Improved
use weighted cross-entropy instead of baselinebaseline · GQA tripletoverall triplet accuracy: 14.7 -> 5.3Hurt
add hub10k to the baseline classifierbaseline · VG8K tripletoverall triplet accuracy: 14.2 -> 14.5Improved

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