AI Paper Atlas

Research map / Detection and segmentation / CNN features

Bootstrapped Masked Autoencoders for Vision BERT Pretraining

Xiaoyi Dong, Jianmin Bao, Ting Zhang, Dongdong Chen, Weiming Zhang, Lu Yuan, Dong Chen, Fang Wen, Nenghai Yu

ECCV 2022 · Detection and segmentation › CNN features · approach: Convolution-transformer backbones · 74 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 bootstrapped feature prediction (momentum-encoder target) (ablation)MAE, 300 epochs · ImageNet-1K valtop-1 accuracy: 83.2 -> 83.6Improved
add feature injection context for the regressor (ablation)BootMAE · ImageNet-1K valtop-1 accuracy: 83.6 -> 83.9Improved
add feature injection context for the predictor as well (ablation)BootMAE · ImageNet-1K valtop-1 accuracy: 83.9 -> 84.0Improved
use block-wise instead of random masking with feature-level target (ablation)MAE 300 epochs · ImageNet-1K valtop-1 accuracy: 83.1 -> 83.6Improved
use block-wise instead of random masking with pixel-level target (ablation)MAE 300 epochs · ImageNet-1K valtop-1 accuracy: 83.2 -> 82.8Hurt
increase regressor/predictor depth from 1 to 2 (ablation)BootMAE · ImageNet-1K valtop-1 accuracy: 83.4 -> 84.0Improved
increase depth from 2 to 8 (ablation)BootMAE · ImageNet-1K valtop-1 accuracy: 84.0 -> 84.0No clear effect
increase regressor dimension from 256 to 512 (ablation)BootMAE · ImageNet-1K valtop-1 accuracy: 83.8 -> 84.0Improved
feed 100% rather than 25% of the image to the momentum encoder (ablation)BootMAE · ImageNet-1K valtop-1 accuracy: 83.8 -> 84.0Improved
feed 100% rather than 25% of image to momentum encoder: training iteration time (ablation)BootMAE · ImageNet-1K valiteration time (s): 0.407 -> 0.660Hurt
use low-level instead of high-level context for pixel regressor (ablation)BootMAE · ImageNet-1K valtop-1 accuracy: 83.4 -> 84.0Improved
use low-level context for pixel regressor instead of middle-level (predictor high-level) (ablation)BootMAE · ImageNet-1K valtop-1 accuracy: 83.7 -> 84.0Improved

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