Research map / Graph learning / Knowledge and temporal graphs
Pixels to Graphs by Associative Embedding
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use end-to-end graph prediction from image pixels without RPN proposals | Our model · Visual Genome SGGen | R@50: 3.4 -> 6.7 | Helped |
| Use end-to-end graph prediction from image pixels without RPN proposals | Our model · Visual Genome SGGen | R@100: 4.2 -> 7.8 | Helped |
| Use the proposed graph prediction method | Visual Genome SGGen (w/ RPN) · Visual Genome SGGen (w/ RPN) | R@50: 3.4 -> 9.7 | Helped |
| Use the proposed graph prediction method | Visual Genome SGGen (w/ RPN) · Visual Genome SGGen (w/ RPN) | R@100: 4.2 -> 11.3 | Helped |
| Use the proposed graph prediction method | Visual Genome SGCls · Visual Genome SGCls | R@50: 21.7 -> 26.5 | Helped |
| Use the proposed graph prediction method | Visual Genome SGCls · Visual Genome SGCls | R@100: 24.4 -> 30.0 | Helped |
| Use the proposed graph prediction method | Visual Genome PredCls · Visual Genome PredCls | R@50: 44.8 -> 68.0 | Helped |
| Use the proposed graph prediction method | Visual Genome PredCls · Visual Genome PredCls | R@100: 53.0 -> 75.2 | Helped |
| Train longer with more efficient code | Our model · Visual Genome SGGen (no RPN) | R@50: 6.7 -> 15.5 | Helped |
| Train longer with more efficient code | Our model · Visual Genome SGGen (no RPN) | R@100: 7.8 -> 18.8 | Helped |
| Train longer with more efficient code | Our model · Visual Genome SGCls | R@50: 26.5 -> 35.7 | Helped |
| Train longer with more efficient code | Our model · Visual Genome SGCls | R@100: 30.0 -> 38.4 | Helped |
Closest papers by meaning
- LinkNet: Relational Embedding for Scene Graph (NeurIPS 2018 · similarity 0.91)
- Image Generation From Scene Graphs (CVPR 2018 · similarity 0.90)
- Generating Triples With Adversarial Networks for Scene Graph Construction (AAAI 2018 · similarity 0.90)
- Scene Graph Generation by Iterative Message Passing (CVPR 2017 · similarity 0.89)
- Learning Convolutional Neural Networks for Graphs (ICML 2016 · similarity 0.89)
- Learning Physical Graph Representations from Visual Scenes (NeurIPS 2020 · similarity 0.88)
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.