AI Paper Atlas

Research map / Detection and segmentation / CNN features

Decoupled DETR: Spatially Disentangling Localization and Classification for Improved End-to-End Object Detection

Manyuan Zhang, Guanglu Song, Yu Liu, Hongsheng Li

ICCV 2023 · Detection and segmentation › CNN features · approach: Region-based and query-based detectors · 41 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 disentangled feature learning (split cross-attention, shared self-attention) (ablation)Conditional DETR + mini-detector · COCO 2017 valAP: 41.7 -> 43.1Improved
Add task-aware query generation (ablation)SD-DETR with DFL · COCO 2017 valAP: 43.1 -> 43.6Improved
Add task alignment loss (ablation)SD-DETR with DFL and TAQ · COCO 2017 valAP: 43.6 -> 45.5Improved
Fully split decoder versus split cross-attention only (ablation)Conditional DETR + mini-detector · COCO 2017 valAP: 42.2 -> 43.1Improved
Fully copy the decoder to decouple branches (ablation)Conditional DETR + mini-detector · COCO 2017 valAP: 41.7 -> 42.2Improved
Apply SD-DETR design to Conditional DETRConditional-DETR-R50 · COCO 2017 val (50 epochs)AP: 41.0 -> 45.5Improved
Apply SD-DETR design with DC5 backbone against DAB-DETRDAB-DETR-R50-DC5 · COCO 2017 val (50 epochs)AP: 44.5 -> 47.0Improved

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