AI Paper Atlas

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

Towards Understanding Why Mask Reconstruction Pretraining Helps in Downstream Tasks

Jiachun Pan, Pan Zhou, Shuicheng YAN

ICLR 2023 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Self-supervised speech and brain decoding

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
Mask-reconstruction pretraining (SimMIM) instead of supervised training on ResNet50ResNet50 · ImageNet classificationTop-1 accuracy (%): +1.8Improved
Mask-reconstruction pretraining instead of supervised training on ResNet50ResNet50 · VOC07+12 detectionAP75: +4.4Improved
Mask-reconstruction pretraining instead of supervised training on ResNet50ResNet50 · VOC07 transfer classificationTop-1 accuracy (%): +3.6Improved
Mask-reconstruction pretraining instead of supervised training on ResNet50ResNet50 · ADE20K semantic segmentationmIoU: +1.9Improved
Using a teacher network (semantic feature targets) rather than pixel targets in MRP theory analysisTeacher-Student MRP · theoretical analysisfeature capture guaranteeNo 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