Research map / Vision-language and multimodal / Cross-modal learning
Multi-attention Recurrent Network for Human Communication Comprehension
Read the paperCheck 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 the Multi-attention Block for continuous cross-modal dynamics modeling (ablation) | MARN · CMU-MOSI test | A²: 76.5 -> 77.1 | Helped |
| add explicit attention coefficients to outline cross-view dynamics (ablation) | MARN · CMU-MOSI test | A²: 59.3 -> 77.1 | Helped |
| add the Multi-attention Block for continuous cross-modal dynamics modeling (ablation) | MARN · MOUD test | A²: 75.5 -> 81.1 | Helped |
| add explicit attention coefficients to outline cross-view dynamics (ablation) | MARN · MOUD test | F1: 61.9 -> 81.2 | Helped |
| add multiple attentions instead of using only one (ablation) | MARN · CMU-MOSI | A² | Helped |
| increase parameter count in the one-attention model (ablation) | MARN with one attention · CMU-MOSI | performance | No clear effect |
| use different numbers of attention units across tasks and datasets (ablation) | MARN · different tasks and datasets | performance | Mixed |
| use MARN rather than TFN | multimodal sentiment analysis · CMU-MOSI test | A²: 74.6 -> 77.1 | Helped |
| use MARN rather than the strongest listed baseline | multimodal sentiment analysis · ICT-MMMO test | A²: 81.3 -> 86.3 | Helped |
| use MARN rather than the strongest listed baseline | multimodal sentiment analysis · YouTube test | A³: 50.2 -> 54.2 | Helped |
| use MARN rather than the strongest listed baseline | multimodal sentiment analysis · MOUD test | A²: 74.0 -> 81.1 | Helped |
| use MARN rather than the strongest listed baseline | emotion recognition · IEMOCAP emotions | A⁹: 36.0 -> 37.0 | Helped |
Closest papers by meaning
- MODA: MOdular Duplex Attention for Multimodal Perception, Cognition, and Emotion Understanding (ICML 2025 · similarity 0.86)
- Debiased Multimodal Understanding for Human Language Sequences (AAAI 2025 · similarity 0.86)
- Revealing Vision-Language Integration in the Brain with Multimodal Networks (ICML 2024 · similarity 0.84)
- BIG-FUSION: Brain-Inspired Global-Local Context Fusion Framework for Multimodal Emotion Recognition in Conversations (AAAI 2025 · similarity 0.84)
- Multi-to-Single: Reducing Multimodal Dependency in Emotion Recognition Through Contrastive Learning (AAAI 2025 · similarity 0.84)
- AffectGPT: A New Dataset, Model, and Benchmark for Emotion Understanding with Multimodal Large Language Models (ICML 2025 · similarity 0.84)
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.