Research map / Vision-language and multimodal / Robot action and navigation
Vision-Dialog Navigation by Exploring Cross-Modal Memory
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 |
|---|---|---|---|
| add collaborative language, visual, and cross-modal memory | CMN · CVDN Val Seen, Val Unseen, and Test Unseen; Oracle, Navigator, and Mixed supervision | Goal Progress (m): 4.48 -> 5.47; 5.67 -> 6.14; 5.92 -> 7.05; 1.23 -> 2.68; 1.98 -> 2.28; 2.10 -> 2.97; 1.25 -> 2.69; 2.11 -> 2.26; 2.35 -> | Helped |
| add V-mem to recall previous visual/navigation memory (ablation) | CMN · CVDN Val Seen | GP (m): 6.33 -> 7.05 | Helped |
| add V-mem to recall previous visual/navigation memory (ablation) | CMN · CVDN Val Unseen | GP (m): 2.52 -> 2.97 | Helped |
| add L-mem to attend relevant historical dialogs (ablation) | CMN · CVDN Val Seen | GP (m): 6.47 -> 7.05 | Helped |
| add L-mem to attend relevant historical dialogs (ablation) | CMN · CVDN Val Unseen | GP (m): 2.64 -> 2.97 | Helped |
| add cross-modal memory from the previous navigation step (ablation) | CMN · CVDN Val Seen | Goal Progress (m): 2.54 -> 2.95 | Helped |
| add visual memory (ablation) | CMN · CVDN Val Seen | Goal Progress (m): 2.74 -> 2.95 | Helped |
| add language memory (ablation) | CMN · CVDN Val Seen | Goal Progress (m): 2.04 -> 2.95 | Helped |
| concatenate an explicit temporal-order embedding with memory features (ablation) | CMN · CVDN | Goal Progress | No clear effect |
Closest papers by meaning
- Structure-Aware Multimodal Sequential Learning for Visual Dialog (AAAI 2024 · similarity 0.88)
- Dual-Latent Memory Routing for Vision-Language Reasoning (ICML 2026 · similarity 0.88)
- VisMem: Latent Vision Memory Unlocks Potential of Vision-Language Models (CVPR 2026 · similarity 0.87)
- Cross-modal Information Flow in Multimodal Large Language Models (CVPR 2025 · similarity 0.87)
- Towards Long-Horizon Vision-Language Navigation: Platform, Benchmark and Method (CVPR 2025 · similarity 0.87)
- Seeing is Understanding: Unlocking Causal Attention into Modality-Mutual Attention for Multimodal LLMs (ICML 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.