Research map / Vision-language and multimodal / Cross-modal learning
PMR: Prototypical Modal Rebalance for Multimodal Learning
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 |
|---|---|---|---|
| Apply PMR to multimodal fusion | CREMA-D fusion models · CREMA-D | Acc: 57.2 -> 61.8 | Helped |
| Apply PMR to multimodal fusion | AVE fusion models · AVE | Acc: 65.1 -> 68.1 | Helped |
| Apply PMR to multimodal fusion | CG-MNIST fusion models · CG-MNIST | Acc: 59.1 -> 78.5 | Helped |
| Add PER to PMR (ablation) | PMR with concatenation fusion · AVE | Acc: 67.2 -> 67.1 | No clear effect |
| Add PER to PMR (ablation) | PMR with concatenation fusion · CREMA-D | Acc: 60.3 -> 61.1 | Helped |
| Use MMTM with PMR | MMTM fusion · CREMA-D | Acc: 51.6 -> 55.1 | Helped |
| Use PMR with concatenation event localization | AVE event localization · AVE | Acc: 72.7 -> 74.3 | Helped |
| Use 1% of training data for prototype calculation (ablation) | PMR with concatenation · AVE | Acc: 65.4 -> 62.1 | Hurt |
| Increase prototype subset from 1% to 10% (ablation) | PMR with concatenation · CREMA-D | Acc: 54.1 -> 61.1 | Helped |
| Increase prototype subset from 1% to 10% (ablation) | PMR with concatenation · AVE | Acc: 62.1 -> 67.1 | Helped |
| Apply PMR with Adam instead of plain Adam training | Adam-trained model · CREMA-D | Acc: 59.8 -> 65.3 | Helped |
| Apply PMR with Adam instead of plain Adam training | Adam-trained model · AVE | Acc: 64.4 -> 66.5 | Helped |
Closest papers by meaning
- Rethinking Multimodal Learning from the Perspective of Mitigating Classification Ability Disproportion (NeurIPS 2025 · similarity 0.90)
- Beyond Modality Fusion: Deep Ensembles for Multimodal Classification (NeurIPS 2026 · similarity 0.90)
- Modality-Aware SAM: Sharpness-Aware-Minimization Driven Gradient Modulation for Harmonized Multimodal Learning (NeurIPS 2025 · similarity 0.90)
- Asymmetric Reinforcing Against Multi-Modal Representation Bias (AAAI 2025 · similarity 0.90)
- Improving Multimodal Learning Balance and Sufficiency through Data Remixing (ICML 2025 · similarity 0.90)
- Multimodal Negative Learning (NeurIPS 2025 · similarity 0.89)
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.