AI Paper Atlas

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

Ranking Info Noise Contrastive Estimation: Boosting Contrastive Learning via Ranked Positives

David T. Hoffmann, Nadine Behrmann, Juergen Gall, Thomas Brox, Mehdi Noroozi

AAAI 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive representation theory · 38 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
Use ranked class and superclass positives with RINCE instead of SCL-outCifar-100 representation learning · Cifar-100 fineaccuracy: 76.70 ± 0.29 -> 77.59 ± 0.21Improved
Use ranked positives with RINCE instead of SCLCifar-100 representation learning · Cifar-100 fineR@1: 74.45 ± 0.39 -> 77.47 ± 0.16Improved
Use ranked positives with RINCE instead of SCL-outImageNet-100 representation learning · ImageNet-100accuracy: 84.18 -> 84.90Improved
Rank frame, shot, and video clips instead of using InfoNCE frame positives onlyUnsupervised video representation learning · HMDBTop 1 accuracy: 41.5 -> 44.9Improved
Rank frame, shot, and video clips instead of using InfoNCE frame positives onlyUnsupervised video representation learning · UCFRetrieval mAP: 0.0688 -> 0.1395Improved
Treat distant video clips as true positives and force all positives to be similar using log-out (ablation)Unsupervised video representation learning · HMDBTop 1 accuracy: 41.5 -> 41.4Hurt
Treat distant video clips as hard negatives rather than positives (ablation)Unsupervised video representation learning · UCFTop 1 accuracy: 74.5 -> 75.2Improved
Use log-in rather than log-out when positives may include noisy distant clips (ablation)Unsupervised video representation learning · HMDBTop 1 accuracy: 41.4 -> 42.6Improved
Set the rank-2 similarity threshold to 0.45 rather than 0.35 (ablation)RINCE-out-in on ImageNet-100 · ImageNet-100 OODAUROC: 60.92 -> 62.92Improved
Use memory bank size 4096 rather than 2048 (ablation)RINCE-out-in on Cifar-100 · Cifar-100 linear evaluationaccuracy: 77.03 -> 77.46Improved
Set rank-2 temperature to 0.225 rather than 0.125 (ablation)RINCE-in on Cifar-100 · Cifar-100 linear evaluationaccuracy: 75.98 -> 77.18Improved
Train for 2000 rather than 1000 epochs (ablation)Supervised contrastive learning and RINCE · Cifar-100accuracyNo clear effect

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