AI Paper Atlas

Research map / Graph learning / Knowledge and temporal graphs

Rethinking Knowledge Graph Propagation for Zero-Shot Learning

Michael Kampffmeyer, Yinbo Chen, Xiaodan Liang, Hao Wang, Yujia Zhang, Eric P. Xing

CVPR 2019 · Graph learning › Knowledge and temporal graphs · approach: Knowledge graph embeddings · 380 citations (OpenAlex)

Check 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
Use dense ancestor/descendant propagation with distance-based weights instead of a shallow GCNZero-shot classifier prediction with ResNet-50 · ImageNet 2-hopsHit@1 accuracy (%): 26.2 -> 26.6Improved
Use dense ancestor/descendant propagation with distance-based weights instead of a shallow GCNZero-shot classifier prediction with ResNet-50 · ImageNet 3-hopsHit@1 accuracy (%): 6.0 -> 6.3Improved
Use dense ancestor/descendant propagation with distance-based weights instead of a shallow GCNZero-shot classifier prediction with ResNet-50 · ImageNet AllHit@1 accuracy (%): 2.8 -> 3.0Improved
Use DGP instead of SGCN when seen classes are included among candidate labelsZero-shot classifier prediction with ResNet-50 · ImageNet 2-hops+1KHit@1 accuracy (%): 11.9 -> 10.3Hurt
Use DGP instead of GCNZZero-shot classification · AWA2 unseen classesTop-1 accuracy (%): 70.7 -> 77.3Improved
Add distance weighting (ablation)DGP · ImageNet 2-hopsHit@1 accuracy (%): 25.4 -> 26.6Improved
Add two-stage finetuning (ablation)DGP · ImageNet 2-hopsHit@1 accuracy (%): 24.6 -> 26.6Improved
Add two-stage finetuning (ablation)SGCN · ImageNet 2-hopsHit@1 accuracy (%): 24.8 -> 26.2Improved
Increase GCN depth from one layer to two (ablation)SGCN without finetuning · ImageNet 2-hopsHit@1 accuracy (%): 24.8 -> 24.2Hurt
Increase GCN depth from one layer to three (ablation)SGCN without finetuning · ImageNet 2-hopsHit@1 accuracy (%): 24.8 -> 23.9Hurt
Separate ancestor and descendant propagation rather than use the full adjacency matrix twice (ablation)DGP · ImageNet 2-hopsHit@1 accuracy (%): 26.0 -> 26.6Improved
Replace standard graph propagation with graph attention (ablation)Single-hidden-layer graph model · ImageNet 2-hopsHit@1 accuracy (%): 24.8 -> 24.1Hurt

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