Research map / Vision-language and multimodal / Visual question answering
Focal Visual-Text Attention for Memex Question Answering
Junwei Liang, Lu Jiang, Liangliang Cao, Yannis Kalantidis, Li-Jia Li, Alexander G. Hauptmann
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 the FVTA kernel tensor instead of cosine similarity (ablation) | FVTA · MemexQA | accuracy: 0.619 -> 0.669 | Improved |
| add intra-sequence dependency (ablation) | FVTA · MemexQA | accuracy: 0.569 -> 0.669 | Improved |
| add cross-sequence interaction (ablation) | FVTA · MemexQA | accuracy: 0.604 -> 0.669 | Improved |
| add question attention (ablation) | FVTA · MemexQA | accuracy: 0.629 -> 0.669 | Improved |
| add photo (visual) input (ablation) | FVTA · MemexQA | accuracy: 0.577 -> 0.669 | Improved |
| use FVTA attention instead of soft attention | FVTA · MemexQA | accuracy: 0.621 -> 0.669 | Improved |
| use FVTA instead of TGIF temporal attention | FVTA · MemexQA | accuracy: 0.630 -> 0.669 | Improved |
| use attention instead of no attention embedding methods | FVTA · MemexQA | accuracy: 0.563 -> 0.669 | Improved |
| use FVTA instead of TGIF temporal on when questions | FVTA · MemexQA (when questions) | accuracy: 0.522 -> 0.476 | Hurt |
| use FVTA attention for evidence localization instead of TGIF | FVTA · MemexQA evidence grounding | HIT@1: 13.28% -> 15.48% | Improved |
| use FVTA instead of RWMN | FVTA · MovieQA val | accuracy: 0.387 -> 0.410 | Improved |
| use FVTA instead of RWMN | FVTA · MovieQA test | accuracy: 0.363 -> 0.373 | Improved |
Closest papers by meaning
- Focal Visual-Text Attention for Visual Question Answering (CVPR 2018 · similarity 0.98)
- An Analysis of Visual Question Answering Algorithms (ICCV 2017 · similarity 0.90)
- Multi-Level Attention Networks for Visual Question Answering (CVPR 2017 · similarity 0.90)
- VTQA: Visual Text Question Answering via Entity Alignment and Cross-Media Reasoning (CVPR 2024 · similarity 0.88)
- FOCUS: Internal MLLM Representations for Efficient Fine-Grained Visual Question Answering (NeurIPS 2025 · similarity 0.88)
- Not Another Text Benchmark: Putting the “Visual" Back in Visual Question Answering for Large Video Models (NeurIPS 2026 · 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.
