AI Paper Atlas

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

What Makes Objects Similar: A Unified Multi-Metric Learning Approach

Han-Jia Ye, De-Chuan Zhan, Xue-Min Si, Yuan Jiang, Zhi-Hua Zhou

NeurIPS 2016 · Detection and segmentation › Open-set and few-shot detection · approach: Object-level representation learning

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
UM2L-ADS multi-metric vs EGO on Facebook circlesUM2L (ADS) · Facebook_348BER: .426 -> .405Improved
UM2L-ADS vs EGOUM2L (ADS) · Facebook_698BER: .392 -> .420Hurt
UM2L-ADS vs EGO on Facebook_3980UM2L (ADS) · Facebook_3980BER: .538 -> .402Improved
UM2L-ADS vs SCA on synthetic feature pattern discoveryUM2L (ADS) · ccdBER: .250 -> .071Improved
UM2L-ADS vs SCA on reutersUM2L (ADS) · reutersBER: .400 -> .398No clear effect
UM2L-ADS vs ITML/LMNN etc. over 10 UCI setsUM2L (ADS) · 10 benchmark datasets, 3NNW/T/L: 8 / 2 / 0 vs EUCLIDImproved
RGS vs ADS configuration of UM2LUM2L (RGS) · 10 benchmark datasets, 3NNW/T/L vs ISD: 7 / 3 / 0 -> 9 / 1 / 0Improved
RGS instead of ADSUM2L · Glass3NN test error: .312 -> .293Improved
RGS instead of ADSUM2L · Autompg3NN test error: .201 -> .225Hurt
UM2L-ADS vs PLMLUM2L (ADS) · Hayes-r3NN test error: .436 -> .276Improved
UM2L-ADS vs MMLMNNUM2L (ADS) · Autompg3NN test error: .256 -> .201Improved
UM2L-ADS vs MMLMNN on GlassUM2L (ADS) · Glass3NN test error: .296 -> .312Hurt

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