AI Paper Atlas

Research map / Detection and segmentation / CNN features

A Simple Single-Scale Vision Transformer for Object Detection and Instance Segmentation

Wuyang Chen, Xianzhi Du, Fan Yang, Lucas Beyer, Xiaohua Zhai, Tsung-Yi Lin, Huizhong Chen, Jing Li, Xiaodan Song, Zhangyang Wang, Denny Zhou

ECCV 2022 · Detection and segmentation › CNN features · approach: Plain and high-resolution ViTs · 43 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
remove spatial downsampling, multi-scale features and doubled channels (plain ViT) (ablation)ViT, 72M params · COCO valbox AP: vanilla ViT performs best across FLOPs rangeImproved
add multi-scale features to spatial downsampling (ablation)ViT · COCO valbox APImproved
use 18 blocks instead of 12 or 40 (ablation)UViT · COCO valbox APImproved
use 896x896 input instead of 1024x1024 or smaller (ablation)UViT · COCO valbox APImproved
use window scale 1/2 instead of global attention in all blocks (ablation)UViT-B · COCO valbox AP: 52.4 -> 52.3 (GFLOPs 2961.9 -> 1298.7)No clear effect
over-shrink early windows to 1/16 (ablation)UViT-B · COCO valbox AP: 52.5 -> 52.0Hurt
use local windows (1/2) in all deep layers instead of global windows at end (ablation)UViT-B · COCO valbox AP: 52.5 -> 52.2Hurt
progressive windows 4^-1 x14, 2^-1 x2, 1 x2 (UViT+) instead of constant 1/2 (ablation)UViT-B · COCO valbox AP: 52.3 -> 52.5 (GFLOPs 1298.7 -> 1160.1)Improved
add self-training with unlabeled ImageNet-1KUViT-B+ · COCO valbox AP: 52.5 -> 53.9Improved
add self-training with unlabeled ImageNet-1KUViT-B+ · COCO valmask AP: 44.8 -> 46.1Improved
use 32 blocks instead of shallower (ablation)UViT · Pascal VOCmIoUImproved

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