Research map / Vision-language and multimodal / Cross-modal learning
Uncertainty-Aware Multi-modal Learning via Cross-Modal Random Network Prediction
Hu Wang, Jianpeng Zhang, Yuanhong Chen, Congbo Ma, Jodie Avery, Louise Hull, Gustavo Carneiro
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 CRNP uncertainty-weighted fusion to the dual-branch 3D UNet (ablation) | 3D UNet · MMWHS validation split | Dice: 0.8706 -> 0.8865 | Improved |
| add self-attention fusion after CRNP (ablation) | CRNP · MMWHS validation split | Dice: 0.8865 -> 0.8958 | Improved |
| add cross-attention fusion after CRNP (ablation) | CRNP · MMWHS validation split | Dice: 0.8865 -> 0.8909 | Improved |
| use residual fusion instead of replace (ablation) | CRNP · MMWHS validation split | Dice: 0.8815 -> 0.8865 | Improved |
| use residual fusion instead of concat (ablation) | CRNP · MMWHS validation split | Dice: 0.8822 -> 0.8865 | Improved |
| ensemble 3 CRNP models | CRNP · MMWHS MR (whole heart) | Dice: 0.8615 -> 0.8758 | Improved |
| ensemble 3 CRNP models | CRNP · MMWHS CT (whole heart) | Dice: 0.9187 -> 0.9193 | Improved |
| ensemble 3 CRNP models | CRNP · BraTS2020 (TC) | Dice: 0.8372 -> 0.8550 | Improved |
| use both modalities (3D UNet) instead of individual MR model | 3D UNet · MMWHS MR (whole heart) | Dice: 0.6729 -> 0.8535 | Improved |
| replace TMC with CRNP | CRNP · Scene15 | accuracy: 0.6774 -> 0.7057 | Improved |
| replace TMC with CRNP | CRNP · Handwritten | AUROC: 0.9997 -> 0.9996 | No clear effect |
| replace 3D UNet with CRNP | CRNP · MMWHS CT (whole heart) | Dice: 0.8957 -> 0.9187 | Improved |
Closest papers by meaning
- Deep Robust Unsupervised Multi-Modal Network (AAAI 2019 · similarity 0.87)
- Multi-modal Vision Pre-training for Medical Image Analysis (CVPR 2025 · similarity 0.85)
- Provable Dynamic Fusion for Low-Quality Multimodal Data (ICML 2023 · similarity 0.84)
- When to Align, When to Predict: A Phase Diagram for Multimodal Learning (NeurIPS 2026 · similarity 0.83)
- Quantifying and Enhancing Multi-modal Robustness with Modality Preference (ICLR 2024 · similarity 0.83)
- Embracing Unimodal Aleatoric Uncertainty for Robust Multimodal Fusion (CVPR 2024 · similarity 0.83)
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.
