Research map / Vision-language and multimodal / Visual question answering
An Analysis of Visual Question Answering Algorithms
Kushal Kafle, Christopher Kanan
Read the paperCheck your idea against 137,217 papers
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 absurd questions to training (ablation) | Q+I · TDIUC test | Arithmetic MPT: 57.03 -> 51.45 | Hurt |
| add absurd questions to training (ablation) | Q+I · TDIUC test | Activity Recognition accuracy: 48 -> 24 | Hurt |
| add absurd questions to training (ablation) | MCB · TDIUC test | Arithmetic MPT: 66.07 -> 65.75 | Mixed |
| add spatial attention | MCB · TDIUC test | Arithmetic MPT: 65.75 -> 67.90 | Improved |
| add spatial attention | MCB · TDIUC test | Color Attributes accuracy: 56.93 -> 68.54 | Improved |
| add spatial attention | MCB · TDIUC test | Absurd accuracy: 83.44 -> 84.82 | Improved |
| add spatial attention | MCB · TDIUC test | Counting accuracy: 50.29 -> 51.01 | Improved |
| use compositional neural module network instead of MCB | MCB · TDIUC test | Arithmetic MPT: 65.75 -> 62.59 | Hurt |
| use NMN instead of MLP | MLP · TDIUC test | Arithmetic MPT: 60.87 -> 62.59 | Improved |
| use MCB instead of MLP | MLP · TDIUC test | Simple Accuracy: 81.07 -> 79.20 | Hurt |
| use MCB instead of MLP | MLP · TDIUC test | Arithmetic MPT: 60.87 -> 65.75 | Improved |
| use RAU instead of MCB-A | MCB-A · TDIUC test | Simple Accuracy: 81.86 -> 84.26 | Improved |
Closest papers by meaning
- Revisiting Visual Question Answering Baselines (ECCV 2016 · similarity 0.95)
- Visual Question Answering as a Meta Learning Task (ECCV 2018 · similarity 0.95)
- Tips and Tricks for Visual Question Answering: Learnings From the 2017 Challenge (CVPR 2018 · similarity 0.95)
- Making the v in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering (CVPR 2017 · similarity 0.95)
- Answer Them All! Toward Universal Visual Question Answering Models (CVPR 2019 · similarity 0.95)
- When Open-Vocabulary Visual Question Answering Meets Causal Adapter: Benchmark and Approach (AAAI 2025 · similarity 0.93)
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.
