Research map / Vision-language and multimodal / Cross-modal learning
SMIL: Multimodal Learning with Severely Missing Modality
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 SMIL instead of GAN with severely missing text modality | CMU-MOSI · 10% text | accuracy (%): 56.5 -> 60.7 | Helped |
| Use SMIL instead of GAN with severely missing text modality | CMU-MOSI · 20% text | F1 Score: 59.1 -> 62.5 | Helped |
| Use SMIL instead of the strongest listed baseline | MM-IMDb · 10% text | F1 Samples: 45.0 -> 49.2 | Helped |
| Use SMIL instead of the strongest listed baseline | MM-IMDb · 20% text | F1 Micro: 51.0 -> 54.6 | Helped |
| Use K-means modality priors rather than directly generating the missing feature (ablation) | SMIL on MM-IMDb · 10% text | F1 Samples: 0.482 -> 0.492 | Helped |
| Add feature regularization (ablation) | SMIL on MM-IMDb · 10% text | F1 Samples: 0.469 -> 0.492 | Helped |
| Use learned Bayesian feature-regularization distributions instead of a fixed Gaussian (ablation) | SMIL on MM-IMDb · 10% text | F1 Samples: 0.475 -> 0.492 | Helped |
| Use stochastic Bayesian reconstruction and feature regularization instead of deterministic variants (ablation) | SMIL on MM-IMDb · 10% text | F1 Samples: 0.474 -> 0.492 | Helped |
Closest papers by meaning
- Retrieval-Augmented Dynamic Prompt Tuning for Incomplete Multimodal Learning (AAAI 2025 · similarity 0.89)
- Mitigating Pervasive Modality Absence Through Multimodal Generalization and Refinement (AAAI 2025 · similarity 0.87)
- Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness (NeurIPS 2026 · similarity 0.87)
- Deep Correlated Prompting for Visual Recognition with Missing Modalities (NeurIPS 2024 · similarity 0.87)
- Modality-Aware SAM: Sharpness-Aware-Minimization Driven Gradient Modulation for Harmonized Multimodal Learning (NeurIPS 2025 · similarity 0.86)
- Knowledge Bridger: Towards Training-Free Missing Modality Completion (CVPR 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.