AI Paper Atlas

Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning

Region Graph Embedding Network for Zero-Shot Learning

Guo-Sen Xie, Li Liu, Fan Zhu, Fang Zhao, Zheng Zhang, Yazhou Yao, Jie Qin, Ling Shao

ECCV 2020 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Generative feature synthesis for zero-shot · 189 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 parts relation reasoning (PRR) branch (ablation)RGEN · CUB ZSLMCA (%): 75.0 -> 76.1Improved
add parts relation reasoning (PRR) branch (ablation)RGEN · AWA2 ZSLMCA (%): 72.5 -> 73.6Improved
add parts relation reasoning (PRR) branch (ablation)RGEN · SUN ZSLMCA (%): 63.4 -> 63.8Improved
add parts relation reasoning (PRR) branch (ablation)RGEN · APY ZSLMCA (%): 43.9 -> 44.4Improved
add parts relation reasoning (PRR) branch (ablation)RGEN · AWA2 GZSLH: 69.7 -> 71.5Improved
add parts relation reasoning (PRR) branch (ablation)RGEN · CUB GZSLH: 64.7 -> 66.1Improved
use two-layer GCN instead of one-layer GCN (ablation)RGEN PRR branch · CUB ZSLMCA (%): 75.2 -> 76.1Improved
use two-layer GCN instead of three-layer GCN (ablation)RGEN PRR branch · CUB ZSLMCA (%): 75.1 -> 76.1Improved
use two-layer GCN instead of one-layer GCN (ablation)RGEN PRR branch · CUB GZSLH: 65.4 -> 66.1Improved
add balance loss in GZSL training (lambda2 > 0) (ablation)RGEN · CUB and AWA2 GZSLH/tsImproved
use small eta2 weight for PRR branch at fusion, (eta1, eta2) = (0.9, 0.1) (ablation)RGEN · CUB and AWA2 ZSLMCA (%)Improved
vary number of parts K from 1 to 15 (ablation)RGEN · CUB and AWA2MCA and HNo clear effect

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