Research map / Vision-language and multimodal / Visual question answering
Making History Matter: History-Advantage Sequence Training for Visual Dialog
Tianhao Yang, Zheng-Jun Zha, Hanwang Zhang
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 FCA and gated history-awareness (history-aware features) (ablation) | HACAN · VisDial v1.0 (val) | MRR: 0.5837 -> 0.6181 | Improved |
| add multi-head ECA (m=16) instead of single head (m=1) (ablation) | HACAN · VisDial v1.0 (val) | MRR: 0.6285 -> 0.6323 | Improved |
| add single-head ECA after FCA (ablation) | HACAN · VisDial v1.0 (val) | MRR: 0.6181 -> 0.6285 | Improved |
| stack attention modules three times residually (ablation) | HACAN · VisDial v1.0 (val) | MRR: 0.6323 -> 0.6391 | Improved |
| add HAST training (ablation) | HACAN · VisDial v1.0 (val) | MRR: 0.6391 -> 0.6445 | Improved |
| add HAST training (ablation) | HACAN · VisDial v1.0 (val) | R@1: 50.44 -> 51.20 | Improved |
| add HAST training (ablation) | HCIAE-M · VisDial v1.0 (val) | MRR: 0.6156 -> 0.6177 | Improved |
| add HAST training (ablation) | HCIAE · VisDial v1.0 (val) | MRR: 0.6227 -> 0.6243 | Improved |
| add HAST training (ablation) | HACAN · GuessWhat?! guesser (test) | test error: 34.1% -> 33.2% | Improved |
| add HAST training (ablation) | HRED · GuessWhat?! guesser (test) | test error: 39.0% -> 38.4% | Improved |
| add HAST training (ablation) | HRED+VGG · GuessWhat?! guesser (test) | test error: 39.6% -> 38.9% | Improved |
| sum only over top-5 negative answers in the advantage (ablation) | HAST · VisDial | training time: saved 95% training time with a slight performance drop | Mixed |
Closest papers by meaning
- Modality-Balanced Models for Visual Dialogue (AAAI 2020 · similarity 0.89)
- UTC: A Unified Transformer With Inter-Task Contrastive Learning for Visual Dialog (CVPR 2022 · similarity 0.88)
- Structure-Aware Multimodal Sequential Learning for Visual Dialog (AAAI 2024 · similarity 0.87)
- Memory-T1: Reinforcement Learning for Temporal Reasoning in Multi-session Agents (ICLR 2026 · similarity 0.86)
- V^2Dial: Unification of Video and Visual Dialog via Multimodal Experts (CVPR 2025 · similarity 0.86)
- TA&AT: Enhancing Task-Oriented Dialog with Turn-Level Auxiliary Tasks and Action-Tree Based Scheduled Sampling (AAAI 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.
