Field dossier · from the corpus

Scene graph

0predicted collisions

Where this field is predicted to collide

No open collisions on this field in the current run.

Most cited work

SuperGlue: Learning Feature Matching With Graph Neural Networks

20202,853 citationsDOI ↗

Neural Motifs: Scene Graph Parsing with Global Context

20181,124 citationsDOI ↗

Exploring Visual Relationship for Image Captioning

2018944 citationsDOI ↗

Unbiased Scene Graph Generation From Biased Training

2020746 citationsDOI ↗

Signals

4

Papers · in corpus

5.7k

Citations · corpus total

0.0%

Recent share · since 2024

2018–2020

Span · publication years

Who is doing the work

Microsoft (United States)US2
Carnegie Mellon UniversityUS1
Art Institute of PortlandUS1
Microsoft (Finland)FI1
Pacific Northwest National LaboratoryUS1
Block Engineering (United States)US1
Allen Institute for Artificial IntelligenceUS1
Allen InstituteUS1

Most published researchers

Rowan Zellers1
Sam Thomson1
Mark Yatskar1
Daniel DeTone1
Tomasz Malisiewicz1
Paul-Edouard Sarlin1

Already lives beside

Fields this one already co-publishes with. A predicted collision is a departure from this neighbourhood, not an extension of it.

Set (abstract data type)2 Graph2 Artificial intelligence2 Baseline (sea)1 Causal inference1 Inference1 Machine learning1 Counterfactual thinking1 Image (mathematics)1 Debiasing1