AI Paper Atlas

Research map / 3D vision / Depth and stereo

In Pursuit of Pixel Supervision for Visual Pre-training

Lihe Yang, Shang-Wen Li, Yang Li, Xinjie Lei, Dong Wang, Abdelrahman Mohamed, Saining Xie, Hengshuang Zhao, Kaiming He, Hu Xu

CVPR 2026 · 3D vision › Depth and stereo · approach: Depth from light field and defocus · 9 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 self-curated MetaCLIP-S instead of uncurated MetaCLIP 2B (ablation)MAE-style encoder · ADE20KmIoU: 44.7 -> 46.8Improved
Pretrain on self-curated MetaCLIP-S instead of uncurated MetaCLIP 2B (ablation)MAE-style encoder · NYUv2 depthRMSE: 0.351 -> 0.321Improved
Use IN-21K instead of IN-1K as pretraining data (ablation)MAE-style encoder · ADE20KmIoU: 42.9 -> 44.8Improved
Increase decoder depth from 8 to 32 (ablation)MAE on IN-21K · IN-1K k-NNaccuracy: 35.3 -> 55.8Improved
Increase decoder depth from 8 to 32 (ablation)MAE on IN-21K · ADE20KmIoU: 35.8 -> 40.4Improved
Make decoder even larger (768x32) (ablation)MAE on IN-21K · dense downstream tasksoverallHurt
Mask 2x2 patch blocks instead of single patches (ablation)MAE (512x8 decoder, 75% mask) · IN-1K k-NNaccuracy: +19.0Improved
Mask 2x2 patch blocks instead of single patches (ablation)MAE (512x8 decoder, 75% mask) · NYUv2 depthRMSE: 0.431 -> 0.362Improved
Increase masking block to 8x8 patches (ablation)MAE · downstream tasksoverallHurt
Increase class tokens from 1 to 4 (ablation)MAE · IN-1K k-NNaccuracy: 63.3 -> 75.1Improved
Increase class tokens from 1 to 4 (ablation)MAE · dense prediction tasksoverallNo clear effect
Combine all three MAE updates (decoder, mask block, CLS tokens) (ablation)MAE on curated 2B images · NYUv2 depthRMSE: 0.392 -> 0.321Improved

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