AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

PASS: An ImageNet replacement for self-supervised pretraining without humans

Yuki M Asano, Christian Rupprecht, Andrew Zisserman, Andrea Vedaldi

NeurIPS 2021 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Pretext-task visual pre-training · 10 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
pretrain on PASS instead of ImageNet-1k (transfer)MoCo-v2 · COCO bbox detectionAP: 38.7 -> 38.0No clear effect
pretrain on PASS instead of ImageNet-1kMoCo-v2 · Linear probing ImageNet-1ktop-1 accuracy: 67.5 -> 59.5Hurt
pretrain on PASS instead of ImageNet-1kMoCo-v2 · Linear probing Herbarium19top-1 accuracy: 42.3 -> 46.5Improved
pretrain on PASS instead of ImageNet-1kMoCo-v2 · Linear probing CIFAR-10top-1 accuracy: 91.6 -> 81.0Hurt
remove images with humans from ImageNet pretraining (ablation)MoCo-v2 · COCO dense poseAP_GPS: 65.0 -> 64.8No clear effect
include humans in PASS (skip human filtering) (ablation)MoCo-v2 on PASS · Linear probing ImageNet-1ktop-1 accuracy: 59.5 -> 59.7No clear effect
match IN-1k class statistics (~IN800 split) (ablation)MoCo-v2 on PASS · Linear probing ImageNet-1ktop-1 accuracy: 59.5 -> 62.4Improved
match IN-1k class statistics (~IN800 split) (ablation)MoCo-v2 on PASS · Linear probing Places205top-1 accuracy: 52.8 -> 51.5Hurt
use all 6M cleaned images (Random6M) instead of 1.28M balanced PASS (ablation)MoCo-v2 · Linear probing ImageNet-1ktop-1 accuracy: 59.5 -> 61.3Improved
use DINO instead of MoCo-v2 on PASS (ablation)PASS pretraining · 4-shot Pascal VOC07mAP: 57.6 -> 64.3Improved
use SwAV instead of MoCo-v2 on PASS (ablation)PASS pretraining · Linear probing Herbarium19top-1 accuracy: 46.5 -> 54.4Improved
lower RRC minimum crop size from 0.2 to 0.1 (ablation)MoCo-v2 on PASS · Linear probing Herbarium19top-1 accuracy: 46.5 -> 48.9Improved

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