Research map / Vision-language and multimodal / Cross-modal learning
Decoupled Multimodal Distilling for Emotion Recognition
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 |
|---|---|---|---|
| Use DMD instead of the strongest listed baseline | CMU-MOSI aligned · CMU-MOSI aligned | ACC7 (%): 40.6 -> 41.4 | Helped |
| Use DMD instead of the strongest listed baseline | CMU-MOSI unaligned · CMU-MOSI unaligned | ACC7 (%): 40.8 -> 41.9 | Helped |
| Use DMD instead of the strongest listed baseline | CMU-MOSEI aligned · CMU-MOSEI aligned | ACC7 (%): 52.5 -> 53.7 | Helped |
| Use DMD instead of the strongest listed baseline | CMU-MOSEI unaligned · CMU-MOSEI unaligned | ACC7 (%): 52.4 -> 54.6 | Helped |
| Add feature decoupling (FD) to the unimodal MOSEI baseline | MOSEI unimodal models · CMU-MOSEI | Mean ACC2 (%): 64.3 -> 70.1 | Helped |
| Add HomoGD to the feature-decoupled model (ablation) | DMD · CMU-MOSI unaligned | ACC7: 34.7 -> 37.2 | Helped |
| Add HomoGD to the feature-decoupled model (ablation) | DMD · CMU-MOSEI unaligned | ACC7: 51.6 -> 52.4 | Helped |
| Add crossmodal attention (CA) to HeteroGD (ablation) | DMD · CMU-MOSI unaligned | ACC7: 37.5 -> 41.9 | Helped |
| Add crossmodal attention (CA) to HeteroGD (ablation) | DMD · CMU-MOSEI unaligned | ACC7: 52.4 -> 54.6 | Helped |
| Add HeteroGD to the model with FD, HomoGD, and CA (ablation) | DMD · CMU-MOSI unaligned | ACC7: 38.8 -> 41.9 | Helped |
| Add HeteroGD to the model with FD, HomoGD, and CA (ablation) | DMD · CMU-MOSEI unaligned | ACC7: 53.2 -> 54.6 | Helped |
| Add feature decoupling (FD) (ablation) | DMD · CMU-MOSI unaligned | ACC7: 32.4 -> 41.9 | Helped |
Closest papers by meaning
- Multi-to-Single: Reducing Multimodal Dependency in Emotion Recognition Through Contrastive Learning (AAAI 2025 · similarity 0.89)
- OV-MER: Towards Open-Vocabulary Multimodal Emotion Recognition (ICML 2025 · similarity 0.89)
- AffectGPT: A New Dataset, Model, and Benchmark for Emotion Understanding with Multimodal Large Language Models (ICML 2025 · similarity 0.88)
- EMOE: Modality-Specific Enhanced Dynamic Emotion Experts (CVPR 2025 · similarity 0.88)
- A Unified Self-Distillation Framework for Multimodal Sentiment Analysis with Uncertain Missing Modalities (AAAI 2024 · similarity 0.88)
- Correlation-Decoupled Knowledge Distillation for Multimodal Sentiment Analysis with Incomplete Modalities (CVPR 2024 · 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.