Research map / Image generation / Text-to-image diffusion
Unbiased Missing-modality Multimodal Learning
Ruiting Dai, Chenxi Li, Yandong Yan, Lisi Mo, Ke Qin, Tao He
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 VP-SDE score model to base recovery (ablation) | MD2N base · CMU-MOSI (30-70% missing) | ACC2: 76.5 -> 76.7 | Improved |
| add VP-SDE score model to base recovery (ablation) | MD2N base · CMU-MOSI (30-70% missing) | F1: 73.4 -> 76.6 | Improved |
| add multi-stage process on top of VP-SDE (ablation) | MD2N · CMU-MOSEI (30-70% missing) | ACC2: 79.4 -> 79.9 | Improved |
| add duplex diffusion on top of multi-stage (ablation) | MD2N · CMU-MOSI (30-70% missing) | ACC2: 77.1 -> 80.0 | Improved |
| add duplex diffusion on top of multi-stage (ablation) | MD2N · CMU-MOSEI (30-70% missing) | ACC2: 79.9 -> 82.6 | Improved |
| add multi-stage duplex diffusion module to MCTN (ablation) | MCTN · CMU-MOSI (30% missing) | ACC2: 71.3 -> 75.7 | Improved |
| add multi-stage duplex diffusion module to MCTN (ablation) | MCTN · CMU-MOSEI (30% missing) | ACC2: 76.9 -> 79.0 | Improved |
| add multimodal feature extractor (ablation) | MD2N · CMU-MOSI (30% missing) | ACC2: 81.2 -> 83.4 | Improved |
| add decoder (ablation) | MD2N · CMU-MOSI (30% missing) | ACC2: 80.4 -> 83.4 | Improved |
| add both feature extractor and decoder (ablation) | MD2N · CMU-MOSEI (30% missing) | ACC2: 79.4 -> 84.3 | Improved |
| shorten refinement stage by decreasing t1 from 200 to 20 (ablation) | MD2N · CMU-MOSEI (30% missing) | ACC2 | Hurt |
| use MD2N instead of IMDer | MD2N · CMU-MOSEI (A only) | ACC2: 63.8 -> 69.7 | Improved |
Closest papers by meaning
- Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness (NeurIPS 2026 · similarity 0.90)
- Mitigating Pervasive Modality Absence Through Multimodal Generalization and Refinement (AAAI 2025 · similarity 0.89)
- RedCore: Relative Advantage Aware Cross-Modal Representation Learning for Missing Modalities with Imbalanced Missing Rates (AAAI 2024 · similarity 0.88)
- Towards General Modality Translation with Contrastive and Predictive Latent Diffusion Bridge (NeurIPS 2025 · similarity 0.88)
- Asymmetric Reinforcing Against Multi-Modal Representation Bias (AAAI 2025 · similarity 0.88)
- Gradient-Guided Modality Decoupling for Missing-Modality Robustness (AAAI 2024 · similarity 0.88)
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.
