AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Attribute Recognition by Joint Recurrent Learning of Context and Correlation

Jingya Wang, Xiatian Zhu, Shaogang Gong, Wei Li

ICCV 2017 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 165 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 encoder LSTM for intra-person attribute context (ablation)JRL · PETAmAP_cls: 83.45 -> 85.67Improved
add encoder LSTM for intra-person attribute context (ablation)JRL · RAPmAP_cls: 75.19 -> 77.81Improved
add inter-person similarity context from top-k exemplars (ablation)JRL · PETAmAP_cls: 85.02 -> 85.67Improved
add inter-person similarity context from top-k exemplars (ablation)JRL · RAPmAP_cls: 76.94 -> 77.81Improved
use ensemble of 10 attribute orders instead of average single-order model (ablation)JRL · PETAmAP_cls: 82.13 -> 85.67Improved
use ensemble of 10 attribute orders instead of average single-order model (ablation)JRL · RAPmAP_cls: 74.74 -> 77.81Improved
add recurrent attribute attention (ablation)JRL · PETAmAP_cls: 84.03 -> 85.67Improved
add recurrent attribute attention (ablation)JRL · RAPmAP_cls: 75.96 -> 77.81Improved
use JRL instead of DeepMAR with 25% of training dataJRL vs DeepMAR · PETAmAP_cls: mAP drop 3.64% vs 6.23% from 100% to 25% dataImproved
use JRL instead of DeepMAR with 25% of training dataJRL vs DeepMAR · RAPmAP_cls: mAP drop 3.55% vs 5.73% from 100% to 25% dataImproved
use JRL instead of ACN on RAP precisionJRL vs ACN · RAPmPrc_ins: 80.12% vs 78.11%Hurt
use JRL instead of SR CNN-RNNJRL vs SR CNN-RNN · PETAmAP_cls: 82.83 -> 85.67Improved

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