AI Paper Atlas

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

Learning of Visual Relations: The Devil Is in the Tails

Alakh Desai, Tz-Ying Wu, Subarna Tripathi, Nuno Vasconcelos

ICCV 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive learning for imbalanced data · 102 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 two-stage decoupled training instead of single-stage (ablation)DT2 · VG150 SGClsmR@100: 12.4 -> 26.7Improved
use class-balanced sampling instead of random sampling in single-stage training (ablation)DT2 single stage · VG150 SGClsmR@100: 11.2 -> 12.4Improved
use separate entity-CBS and predicate-CBS instead of predicate-CBS for both (ablation)DT2 · VG150 SGClsmR@100: 14.3 -> 26.7Improved
add ACBS distillation between P-step and E-step to independent CBS (ablation)DT2 · VG150 SGClsmR@100: 26.7 -> 27.5Improved
add the appearance branch (ablation)DT2-ACBS · VG150 PredClsmR@100: 26.8 -> 39.7Improved
add the appearance branch (ablation)DT2-ACBS · VG150 SGClsmR@100: 16.3 -> 27.5Improved
add the appearance branch (ablation)DT2-ACBS · VG150 SGClssubject-object entity accuracy: 25.77 -> 26.26No clear effect
use P-step as teacher instead of E-step (ablation)DT2-ACBS · VG150 SGClsmR@100: 22.0 -> 27.5Improved
replace MOTIFS with DT2-ACBS using long-tail-aware training and simple architectureDT2-ACBS vs TDE-MOTIFS-SUM · VG150 PredCls tail classesmR@100: 6.1 -> 38.6Improved
DT2-ACBS versus TDE-MOTIFS-SUM on head classesDT2-ACBS · VG150 PredCls head classesmR@100: 44.9 -> 35.1Hurt
DT2-ACBS versus PCPL on PredClsDT2-ACBS · VG150 PredClsmR@100: 37.8 -> 39.7Improved
DT2-ACBS versus PCPL on SGDetDT2-ACBS · VG150 SGDetmR@100: 11.7 -> 24.4Improved

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