AI Paper Atlas

Research map / Detection and segmentation / CNN features

Visible and Clear: Finding Tiny Objects in Difference Map

Bing Cao, Haiyu Yao, Pengfei Zhu, Qinghua Hu

ECCV 2024 · Detection and segmentation › CNN features · approach: Box regression and loss design · 56 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 DGFE module on top of reconstruction head (ablation)Cascade R-CNN · DroneSwarmsAP: 36.5 -> 38.3Improved
add reconstruction head alone (no difference map guidance) (ablation)Cascade R-CNN · DroneSwarmsAP: 36.4 -> 36.5No clear effect
use element-wise attention instead of element-wise multiplication (ablation)DGFE · DroneSwarmsAP: 36.2 -> 38.3Improved
use element-wise attention instead of concatenation (ablation)DGFE · DroneSwarmsAP: 36.6 -> 38.3Improved
add learnable threshold filtration vs none (ablation)DGFE · VisDrone2019AP: 27.0 -> 27.3Improved
make threshold learnable instead of fixed (ablation)DGFE · VisDrone2019AP: 27.1 -> 27.3Improved
use high-frequency difference map instead of pixel difference map (ablation)SR-TOD · DroneSwarmsAP: 38.3 -> 38.4No clear effect
add SR-TOD to Cascade R-CNNCascade R-CNN · DroneSwarmsAP: 36.4 -> 38.3Improved
add SR-TOD to DINODINO · DroneSwarmsAP: 35.4 -> 35.6Improved
add SR-TOD to RFLARFLA · VisDrone2019AP: 27.2 -> 27.8Improved
add SR-TOD to DetectoRSDetectoRS · AI-TODAP: 14.6 -> 24.0Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 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