AI Paper Atlas

Research map / Detection and segmentation / Weakly supervised detection and segmentation

Dense Teacher: Dense Pseudo-Labels for Semi-Supervised Object Detection

Hongyu Zhou, Zheng Ge, Songtao Liu, Weixin Mao, Zeming Li, Haiyan Yu, Jian Sun

ECCV 2022 · Detection and segmentation › Weakly supervised detection and segmentation · approach: Pseudo-labeling for semi-supervised detection · 159 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
replace pseudo-boxes with dense pseudo-labels over all regionsUnbiased Teacher on FCOS · COCO-Standard 10%AP: 31.52 -> 32.0Improved
add FRS-based region division (top k% learning region) (ablation)Dense Teacher · COCO-Standard 10%AP: 32.00 -> 33.34Improved
add unsupervised regression branch (ablation)Dense Teacher · COCO-Standard 10%AP: 33.34 -> 35.11Improved
select hard negatives instead of suppressing them (ablation)Dense Teacher · COCO-Standard 10%AP: 31.56 -> 33.34Improved
select hard negatives instead of ignoring them (ablation)Dense Teacher · COCO-Standard 10%AP: 32.64 -> 33.34Improved
learning region k=1 instead of 0.1 (ablation)Dense Teacher · COCO-Standard 10%AP: 25.76 -> 35.11Improved
learning region k=1 instead of 5 (ablation)Dense Teacher · COCO-Standard 10%AP: 33.85 -> 35.11Improved
unsupervised weight 4 instead of 8 (ablation)Dense Teacher · COCO-Standard 10%AP: 33.81 -> 35.11Improved
unsupervised weight 2 instead of 4 (ablation)Dense Teacher · COCO-FullAP: 43.04 -> 44.92Improved
replace pseudo-boxes with dense labels on anchor-based RetinaNetDense Teacher · COCO (RetinaNet)AP: 28.9 -> 31.1Improved
replace pseudo-boxes with dense labels on CrowdHumanDense Teacher · CrowdHuman (FCOS)mMR: 62.8 -> 60.7Improved
large scale jittering and longer scheduleDense Teacher · COCO-FullmAP: 43.90 -> 44.94 -> 46.12Improved

Later papers that compared with this method

7 later accepted papers report Dense Teacher as a baseline in their results tables; 7 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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