AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Masked Image Residual Learning for Scaling Deeper Vision Transformers

Guoxi Huang, Hongtao Fu, Adrian G. Bors

NeurIPS 2023 · Architectures and efficiency › Attention and transformers · approach: Efficient and robust vision transformers

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
Use masked image residual learning instead of standard masked image reconstructionViT-B-48 pre-training · ImageNet-1Kfine-tuning top-1 accuracy: 85.9 -> 86.2Improved
Use MIRL rather than MAE pre-trainingViT-B transfer to MS-COCO · MS-COCO, Mask R-CNN 1×box AP: 48.4 -> 49.3Improved
Add feature-level loss to MIRL (ablation)ViT-B · ImageNet-1Kfine-tuning accuracy: 83.5 -> 83.6Improved
Add VGG loss to MIRL (ablation)ViT-B · ImageNet-1Kfine-tuning accuracy: 83.5 -> 83.8Improved
Add feature-level loss to MIRL (ablation)ViT-B · ImageNet-1Kpre-training time: 1.0× -> 1.8×Hurt
Add VGG loss to MIRL (ablation)ViT-B · ImageNet-1Kpre-training time: 1.0× -> 2.5×Hurt
Increase encoder segmentation from 2 to 4 segments (ablation)MIRL ViT-B-24 · ImageNet-1Kfine-tuning accuracy: 84.3 -> 84.7Improved
Increase encoder segmentation from 4 to 6 segments (ablation)MIRL ViT-B-24 · ImageNet-1Kfine-tuning accuracy: 84.7 -> 84.6Hurt
Add densely interconnected decoding (DID) (ablation)MIRL ViT-B-24 · ImageNet-1Kfine-tuning accuracy: 84.5 -> 84.7Improved
Add densely interconnected decoding (DID) (ablation)MIRL ViT-B-48 · ImageNet-1Kfine-tuning accuracy: 85.0 -> 85.3Improved
Replace residual-prediction shortcuts with independent decoders that each reconstruct masked content (ablation)segmented ViT pre-training · ImageNet-1Kfine-tuning accuracy: 83.5 -> 83.2Hurt
Replace main/residual targets with coarse/fine image targets (ablation)segmented ViT-B pre-training · ImageNet-1Kfine-tuning accuracy: 83.5 -> 83.1Hurt

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