Research map / Graph learning / Knowledge and temporal graphs
Explicit Knowledge Incorporation for Visual Reasoning
Yifeng Zhang, Ming Jiang, Qi Zhao
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Add KI-Net knowledge incorporation and G-Relate high-order attention | Explicit visual reasoning model · GQA test-dev | overall accuracy: 7.6% | Improved |
| Add KI-Net knowledge incorporation and G-Relate high-order attention | Explicit visual reasoning model · VQAv2 val | overall accuracy: 3.2% | Improved |
| Add G-Relate high-order attention (ablation) | Visual reasoning model without KI-Net (Baseline I) · GQA test-dev | overall accuracy: 56.61 -> 58.74 | Improved |
| Add G-Relate high-order attention (ablation) | Visual reasoning model without KI-Net (Baseline I) · VQAv2 val | overall accuracy: 64.11 -> 65.38 | Improved |
| Add KI-Net external knowledge incorporation (ablation) | Visual reasoning model without G-Relate (Baseline I) · GQA test-dev | overall accuracy: 56.61 -> 60.69 | Improved |
| Add KI-Net external knowledge incorporation (ablation) | Visual reasoning model without G-Relate (Baseline I) · VQAv2 val | overall accuracy: 64.11 -> 65.98 | Improved |
| Explicitly incorporate external entities and predicates into scene graphs | Scene graph generation with KI-Net · VQAv2 val | mR@100: 6.9 -> 7.3 | Improved |
| Use WordNet, ConceptNet, and Visual Genome together rather than no external knowledge (ablation) | KI-Net · VQAv2 val | mR@50: 5.5 -> 6.2 | Improved |
| Add KI-Net knowledge incorporation (ablation) | NSM · VQAv2 val | accuracy: 63.41 -> 63.68 | Improved |
| Add KI-Net knowledge incorporation (ablation) | XNM · VQAv2 val | accuracy: 64.89 -> 65.93 | Improved |
Closest papers by meaning
- Visual Structures Help Visual Reasoning: Addressing the Binding Problem in LVLMs (NeurIPS 2025 · similarity 0.89)
- Seeing as Experts Do: A Knowledge-Augmented Agent for Open-Set Fine-Grained Visual Understanding (CVPR 2026 · similarity 0.89)
- Scene Graph Thinking: Reinforcing Structured Visual Reasoning for Multimodal Large Language Models (ICML 2026 · similarity 0.88)
- VisRes Bench: On Evaluating the Visual Reasoning Capabilities of VLMs (CVPR 2026 · similarity 0.88)
- Semantic-Enriched Latent Visual Reasoning (ICML 2026 · similarity 0.88)
- SLVR: Structured Latent Visual Reasoning via Human-like Reasoning Flows (NeurIPS 2026 · similarity 0.87)
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.
