AI Paper Atlas

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

SLIP: Self-Supervision Meets Language-Image Pre-training

Norman Mu, Alexander Kirillov, David Wagner, Saining Xie

ECCV 2022 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Invariance-based self-supervised learning · 392 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 SimCLR self-supervision to CLIPCLIP ViT-B/16 · ImageNet zero-shottop-1 accuracy (%): 37.6 -> 42.8Improved
add SimCLR self-supervision to CLIPCLIP ViT-S/16 · ImageNet zero-shottop-1 accuracy (%): 32.7 -> 38.3Improved
add SimCLR self-supervision to CLIPCLIP ViT-L/16 · ImageNet zero-shottop-1 accuracy (%): 40.4 -> 46.2Improved
add CLIP language supervision to SimCLRSimCLR ViT-B/16 · ImageNet lineartop-1 accuracy (%): 64.0 -> 72.1Improved
add CLIP language supervision to SimCLR: finetuningSimCLR ViT-B/16 · ImageNet finetuningtop-1 accuracy (%): 82.5 -> 82.6No clear effect
pre-train on YFCC15M instead of ImageNet (ablation)SimCLR ViT-B/16 · ImageNet lineartop-1 accuracy (%): 74.5 -> 64.0Hurt
train for 100 epochs instead of 25 (ablation)SLIP ViT-B/16 · ImageNet zero-shottop-1 accuracy (%): 42.8 -> 45.0Improved
scale model from ViT-B to ViT-L (ablation)SLIP · ImageNet zero-shottop-1 accuracy (%): 45.0 -> 47.9Improved
use multi-task training instead of SimCLR then CLIP finetuning (ablation)SLIP · ImageNet zero-shottop-1 accuracy (%): 41.1 -> 42.8Improved
add color + blur augmentation to CLIP (ablation)CLIP ViT-B/16 · ImageNet zero-shottop-1 accuracy (%): 37.6 -> 38.4Improved
add resize crop + flip augmentation to CLIP (ablation)CLIP ViT-B/16 · ImageNet zero-shottop-1 accuracy (%): 37.6 -> 36.0Hurt
decouple self-supervision and text-supervision (ablation)SLIP ViT-B/16 · ImageNet zero-shottop-1 accuracy (%): 42.8 -> 42.7No 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