Research map / Vision-language and multimodal / Cross-modal learning
Trustworthy Multimodal Regression with Mixture of Normal-inverse Gamma Distributions
Huan Ma, Zongbo Han, Changqing Zhang, Huazhu Fu, Joey Tianyi Zhou, Qinghua Hu
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 |
|---|---|---|---|
| fuse modality NIGs with NIG summation instead of simple average (ablation) | MoNIG · Superconductivity | RMSE: 13.01 -> 12.19 | Improved |
| fuse modality NIGs with NIG summation instead of simple average (ablation) | MoNIG · CT slices | RMSE: 2.87 -> 0.91 | Improved |
| use NIG summation instead of AU-weighted average (ablation) | MoNIG · CT slices | RMSE: 5.71 -> 0.91 | Improved |
| use NIG summation instead of EU-weighted average (ablation) | MoNIG · Superconductivity | RMSE: 13.03 -> 12.19 | Improved |
| add pseudo (feature-concatenation) modality (ablation) | MoNIG · Superconductivity | RMSE: 12.19 -> 11.70 | Improved |
| add pseudo modality (ablation) | MoNIG · CT slices | RMSE: 0.91 -> 0.79 | Improved |
| add pseudo modality (ablation) | MoNIG · CMU-MOSEI unaligned | MAE: 0.598 -> 0.612 | Hurt |
| add pseudo modality (ablation) | MoNIG · CMU-MOSI word aligned | Acc2: 80.2 -> 80.6 | Improved |
| modality-specific mixture of NIGs instead of evidential model on concatenated features under modality noise | evidential regression · Superconductivity eps=0.1 | RMSE: 44.80 -> 21.82 | Improved |
| MoNIG instead of evidential regression on concatenated features under modality noise | evidential regression · CT eps=0.1 | RMSE: 5.35 -> 2.23 | Improved |
| MoNIG instead of evidential regression for uncertainty quality | evidential regression · Superconductivity | UEIR (AU): 13.73 -> 12.65 | Improved |
| MoNIG instead of Gaussian/evidential baselines on sentiment | multimodal sentiment · CMU-MOSI word aligned | MAE: 0.909 -> 0.951 | Mixed |
Closest papers by meaning
- Efficient Analytic Uncertainty Quantification for Multimodal Regression (NeurIPS 2026 · similarity 0.84)
- Bridging the Gap for Test-Time Multimodal Sentiment Analysis (AAAI 2025 · similarity 0.83)
- Towards Robust Multimodal Sentiment Analysis with Incomplete Data (NeurIPS 2024 · similarity 0.83)
- Predictive Dynamic Fusion (ICML 2024 · similarity 0.83)
- A Multi-Focus-Driven Multi-Branch Network for Robust Multimodal Sentiment Analysis (AAAI 2025 · similarity 0.82)
- Toward Robust Incomplete Multimodal Sentiment Analysis via Hierarchical Representation Learning (NeurIPS 2024 · similarity 0.82)
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.
