AI Paper Atlas

Research map / Detection and segmentation / CNN features

Exploring Plain Vision Transformer Backbones for Object Detection

Yanghao Li, Hanzi Mao, Ross Girshick, Kaiming He

ECCV 2022 · Detection and segmentation › CNN features · approach: Plain and high-resolution ViTs · 1,044 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 simple feature pyramid (ablation)ViTDet ViT-L · COCO valAP box: 51.2 -> 54.6Improved
replace simple pyramid with FPN 4-stage (ablation)ViTDet ViT-L · COCO valAP box: 54.6 -> 54.4No clear effect
replace simple pyramid with average-pooling-only pyramid (ablation)ViTDet ViT-L · COCO valAP box: 54.6 -> 54.5No clear effect
add 4 global propagation blocks to window attention (ablation)ViTDet ViT-L · COCO valAP box: 52.9 -> 54.6Improved
add 4 convolutional propagation blocks (ablation)ViTDet ViT-L · COCO valAP box: 52.9 -> 54.8Improved
use shifted windows instead of propagation blocks (ablation)ViTDet ViT-L · COCO valAP box: 54.6 -> 54.0Hurt
place propagation in first 4 blocks only (ablation)ViTDet ViT-L · COCO valAP box: 54.6 -> 52.9Hurt
place propagation in last 4 blocks instead of evenly (ablation)ViTDet ViT-L · COCO valAP box: 54.6 -> 54.3No clear effect
use 24 global blocks instead of 4 (ablation)ViTDet ViT-L · COCO valAP box: 54.6 -> 55.1No clear effect
use MAE IN-1K pre-training instead of supervised IN-1K (ablation)ViTDet ViT-L · COCO valAP box: 49.6 -> 54.6Improved
use IN-21K supervised instead of IN-1K supervised pre-training (ablation)ViTDet ViT-L · COCO valAP box: 49.6 -> 50.6No clear effect
use IN-1K supervised pre-training instead of random init (ablation)ViTDet ViT-L · COCO valAP box: 50.0 -> 49.6Hurt

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