AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Attribute Descent: Simulating Object-Centric Datasets on the Content Level and Beyond

Yue Yao, Liang Zheng, Xiaodong Yang, Milind Napthade, Tom Gedeon

TPAMI 2023 · Detection and segmentation › Open-set and few-shot detection · approach: Object-level representation learning · 19 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
optimize attributes with attribute descent instead of random attributesObjectX synthetic data · VisDA target (synthetic only)per-class accuracy: 65.0 -> 71.1Improved
optimize attributes with attribute descent instead of random attributesPersonX synthetic data · Market (synthetic only)Rank-1: +7.69%Improved
optimize attributes with attribute descent instead of random attributesVehicleX synthetic data · VeRi (synthetic only)Rank-1: 43.56 -> 51.25Improved
augment real data with attribute-descent synthetic data instead of random-attribute dataResNet-50 · VisDA target augmentationper-class accuracy: 91.8 -> 93.6Improved
augment real data with random-attribute synthetic dataIDE · DukemAP: 58.93 -> 55.77Hurt
augment real data with attribute-descent synthetic dataIDE · DukemAP: 58.93 -> 59.10Improved
augment with attribute-descent synthetic data in 1:1 splitResNet-50 · VisDA target (1:1)per-class accuracy: 94.9 -> 95.5Improved
augment with random-attribute synthetic data in 1:1 splitResNet-50 · VisDA target (1:1)per-class accuracy: 94.9 -> 92.0Hurt
add orientation attribute optimization (ablation)attribute descent · VehicleID synthetic onlymAP: -10.52% mAP when omittedImproved
add camera pose attribute optimization (ablation)attribute descent · VehicleID synthetic onlymAP: -3.54% mAP when omittedImproved
add lighting attribute optimization (ablation)attribute descent · VehicleID synthetic onlymAP: -1.6% mAP when omittedImproved
add orientation attribute optimization (ablation)attribute descent · classification and person re-IDtop-1 / mAP: 15.7% and 9.95% drop when omittedImproved

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