Research map / Vision-language and multimodal / Visual question answering
MUREL: Multimodal Relational Reasoning for Visual Question Answering
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 |
|---|---|---|---|
| replace attention with MuRel relational reasoning | VQA model using Bottom-up features · VQA 2.0 val | accuracy: 63.44 -> 65.14 | Helped |
| replace attention with MuRel relational reasoning | VQA model using Bottom-up features · VQA-CP v2 | accuracy: 38.04 -> 39.54 | Helped |
| replace attention with MuRel relational reasoning | VQA model using Bottom-up features · TDIUC | accuracy: 86.96 -> 88.20 | Helped |
| add pairwise modeling (ablation) | MuRel without pairwise modeling or iteration · VQA 2.0 val | accuracy: +0.44 | Helped |
| add pairwise modeling (ablation) | MuRel without pairwise modeling or iteration · VQA-CP v2 | accuracy: +0.24 | Helped |
| add pairwise modeling (ablation) | MuRel without pairwise modeling or iteration · TDIUC | accuracy: +0.36 | Helped |
| add iterative processing (ablation) | MuRel without pairwise modeling or iteration · VQA 2.0 val | accuracy: +0.59 | Helped |
| add iterative processing (ablation) | MuRel without pairwise modeling or iteration · VQA-CP v2 | accuracy: +0.49 | Helped |
| add iterative processing (ablation) | MuRel without pairwise modeling or iteration · TDIUC | accuracy: +0.42 | Helped |
| add pairwise modeling and iterative processing together (ablation) | MuRel without pairwise modeling or iteration · VQA 2.0 val | accuracy: +1.01 | Helped |
| increase reasoning steps from one to two (ablation) | MuRel · VQA 2.0 val | overall accuracy: +0.30 | Helped |
| increase reasoning steps from one to three (ablation) | MuRel · VQA 2.0 val | overall accuracy: +0.57 | Helped |
Closest papers by meaning
- Object Attribute Matters in Visual Question Answering (AAAI 2024 · similarity 0.89)
- SLVR: Structured Latent Visual Reasoning via Human-like Reasoning Flows (NeurIPS 2026 · similarity 0.89)
- FOCUS: Internal MLLM Representations for Efficient Fine-Grained Visual Question Answering (NeurIPS 2025 · similarity 0.89)
- Visual Graph Arena: Evaluating Visual Conceptualization of Vision and Multimodal Large Language Models (ICML 2025 · similarity 0.89)
- Latent Visual Reasoning (ICLR 2026 · similarity 0.89)
- Multimodal Commonsense Knowledge Distillation for Visual Question Answering (Student Abstract) (AAAI 2025 · similarity 0.89)
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.