Research map / Vision-language and multimodal / Visual question answering
Recursive Visual Attention in Visual Dialog
Check 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 |
|---|---|---|---|
| Add recursive visual attention (ablation) | RvA · VisDial v0.9 validation, discriminative | MRR: 0.6551 -> 0.6634 | Helped |
| Add the region proposal network instead of VGG-16 spatial grid features (ablation) | RvA · VisDial v0.9 validation, discriminative | MRR: 0.6436 -> 0.6634 | Helped |
| Add the visual feature filter (ablation) | RvA · VisDial v0.9 validation, discriminative | MRR: 0.6598 -> 0.6634 | Helped |
| Add recursive visual attention (ablation) | RvA · VisDial v1.0 test-standard, discriminative | MRR: 0.6226 -> 0.6303 | Helped |
| Add the region proposal network instead of VGG-16 spatial grid features (ablation) | RvA · VisDial v1.0 test-standard, discriminative | MRR: 0.6060 -> 0.6303 | Helped |
| Add the visual feature filter (ablation) | RvA · VisDial v1.0 test-standard, discriminative | MRR: 0.6294 -> 0.6303 | Helped |
| Add recursive visual attention (ablation) | RvA · VisDial v0.9 validation, generative | MRR: 0.5417 -> 0.5543 | Helped |
| Use RPN region proposals rather than VGG-16 spatial grid features (ablation) | RvA · VisDial v0.9 validation, generative | MRR: 0.5417 -> 0.5543 | Helped |
| Add recursive visual attention (ablation) | RvA with RPN and Bi-LSTM · VisDial v0.9 validation, discriminative | MRR: 0.6551 -> 0.6634 | Helped |
| Use bidirectional LSTM rather than vanilla LSTM (ablation) | RvA with RPN and recursive visual attention · VisDial v0.9 validation, discriminative | MRR: 0.6626 -> 0.6634 | Helped |
| Add recursive visual attention (ablation) | RvA without RPN or Bi-LSTM · VisDial v0.9 validation, discriminative | MRR: 0.6377 -> 0.6418 | Helped |
Closest papers by meaning
- Visual Reference Resolution using Attention Memory for Visual Dialog (NeurIPS 2017 · similarity 0.91)
- Visual Dialog (CVPR 2017 · similarity 0.89)
- Image-Question-Answer Synergistic Network for Visual Dialog (CVPR 2019 · similarity 0.88)
- Factor Graph Attention (CVPR 2019 · similarity 0.88)
- Visual Chain-of-Thought Prompting for Knowledge-Based Visual Reasoning (AAAI 2024 · similarity 0.86)
- Rephrase, Augment, Reason: Visual Grounding of Questions for Vision-Language Models (ICLR 2024 · similarity 0.86)
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.