atlas

Research map / Video / Action recognition

I Know the Relationships: Zero-Shot Action Recognition via Two-Stream Graph Convolutional Networks and Knowledge Graphs

AAAI 2019 · Video › Action recognition · approach: Disentangled action recognition · 223 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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
TS-GCN vs prior SOTAZSAR · UCF101 (inductive)accuracy: 30.3 -> 34.2Helped
TS-GCN vs ZSECOCZSAR · Olympic Sports (inductive)accuracy: 59.8 -> 56.5Hurt
add FV visual features to TS-GCNTS-GCN · UCF101 (inductive)accuracy: +2.1%Helped
two-stream instead of one-stream GCN (ablation)TS-GCN · UCF101accuracy: 25.2 -> 42.8Helped
add self-attention over video segments (ablation)TS-GCN · UCF101accuracy: 39.6 -> 42.8Helped
YFCC100M word embeddings instead of GoogleNews (ablation)TS-GCN · Olympic/HMDB51/UCF101accuracy: 60.5 -> 59.2 (Olympic); 40.5 -> 42.8 (UCF101)Mixed
GCN depth beyond 3 layers (ablation)TS-GCN · three ZSAR benchmarksaccuracyNo clear effect
moderate number of objects K per action (ablation)TS-GCN · three ZSAR benchmarksaccuracyMixed

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