Research map / Label-efficient and robust learning / Knowledge distillation
MMANet: Margin-Aware Distillation and Modality-Aware Regularization for Incomplete 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 |
|---|---|---|---|
| Replace the strongest baseline SF with MMANet | Incomplete multimodal classification · CASIA-SURF | ACER (average): 5.80 -> 3.94 | Helped |
| Replace mmFormer with MMANet | Incomplete multimodal classification · CeFA | ACER (average): 31.52 -> 27.94 | Helped |
| Replace the strongest baseline ESANet with MMANet | Incomplete multimodal segmentation · NYUv2 | mIOU (average): 44.65 -> 45.58 | Helped |
| Replace the strongest baseline ESANet with MMANet | Incomplete multimodal segmentation · Cityscapes | mIOU (average): 71.77 -> 72.20 | Helped |
| Add teacher-guided sample-relation distillation (SP) to SF-MD (ablation) | SF-MD · CASIA-SURF | ACER (average): 7.5 -> 5.9 | Helped |
| Replace SP with uncertainty-weighted margin-aware distillation (MAD) (ablation) | SF-MD · CASIA-SURF | ACER (average): 5.9 -> 4.57 | Helped |
| Add MAD to SF-MD (ablation) | SF-MD · CASIA-SURF | ACER (average): 7.5 -> 4.57 | Helped |
| Add conventional single-modality regularization (SR) (ablation) | SF-MAD · CASIA-SURF | ACER (average): 4.57 -> 4.33 | Helped |
| Replace SR with modality-aware regularization (MAR) (ablation) | SF-MAD · CASIA-SURF | ACER (average): 4.33 -> 3.94 | Helped |
| Add MAR to SF-MAD (ablation) | SF-MAD · CASIA-SURF | ACER (average): 4.57 -> 3.94 | Helped |
| Add SR rather than MAR (ablation) | SF-MAD on the Depth-only combination · CASIA-SURF Depth | ACER: 2.54 -> 1.89 | Mixed |
| Add SR rather than MAR (ablation) | SF-MAD on the RGB+Depth combination · CASIA-SURF RGB+Depth | ACER: 1.23 -> 1.66 | Mixed |
Closest papers by meaning
- Modality-Aware SAM: Sharpness-Aware-Minimization Driven Gradient Modulation for Harmonized Multimodal Learning (NeurIPS 2025 · similarity 0.88)
- CyIN: Cyclic Informative Latent Space for Bridging Complete and Incomplete Multimodal Learning (NeurIPS 2025 · similarity 0.87)
- Balancing Multimodal Training Through Game-Theoretic Regularization (NeurIPS 2025 · similarity 0.86)
- Beyond Modality Fusion: Deep Ensembles for Multimodal Classification (NeurIPS 2026 · similarity 0.86)
- Gradient-Guided Modality Decoupling for Missing-Modality Robustness (AAAI 2024 · similarity 0.86)
- Asymmetric Reinforcing Against Multi-Modal Representation Bias (AAAI 2025 · 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.