Research map / Vision-language and multimodal / Cross-modal learning
SPR: A Structured Prompt Refinement Network for Modality Missing
Hao Chen, Diwei Su, Zhuo Wang, Zuwang He, Menglu Chen, Xiuxing Li, Xia Wu
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 the GIF module to the prompt-only baseline (ablation) | prompt-only baseline · Hateful Memes (70% missing, both modalities) | AUROC: 64.85 -> 66.14 | Improved |
| add the LFR module to the prompt-only baseline (ablation) | prompt-only baseline · Hateful Memes (70% missing, both modalities) | AUROC: 64.85 -> 67.45 | Improved |
| add the CFS module to the prompt-only baseline (ablation) | prompt-only baseline · Hateful Memes (70% missing, both modalities) | AUROC: 64.85 -> 65.73 | Improved |
| add GIF to LFR+CFS (ablation) | SPR · Hateful Memes (70% missing, both modalities) | AUROC: 68.43 -> 71.08 | Improved |
| add LFR to GIF+CFS (ablation) | SPR · Hateful Memes (70% missing, both modalities) | AUROC: 68.56 -> 71.08 | Improved |
| add CFS to GIF+LFR (ablation) | SPR · Hateful Memes (70% missing, both modalities) | AUROC: 67.92 -> 71.08 | Improved |
| add CFS to GIF+LFR (ablation) | SPR · MM-IMDb (70% missing, both modalities) | F1-Macro: 53.82 -> 54.95 | Improved |
| add GIF to LFR+CFS (ablation) | SPR · Food101 (70% missing, both modalities) | accuracy: 83.19 -> 83.92 | Improved |
| use the GIF module instead of attention (ablation) | SPR · Hateful Memes (70% missing, both modalities) | AUROC: 63.95 -> 66.14 | Improved |
| use the GIF module instead of MLP (ablation) | SPR · Hateful Memes (70% missing, both modalities) | AUROC: 65.12 -> 66.14 | Improved |
| add residual connection in GIF (ablation) | SPR · Hateful Memes (70% missing, both modalities) | AUROC: 63.13 -> 66.14 | Improved |
| add residual connection in LFR (ablation) | SPR · Hateful Memes (70% missing, both modalities) | AUROC: 63.27 -> 67.45 | Improved |
Closest papers by meaning
- Deep Correlated Prompting for Visual Recognition with Missing Modalities (NeurIPS 2024 · similarity 0.91)
- Retrieval-Augmented Dynamic Prompt Tuning for Incomplete Multimodal Learning (AAAI 2025 · similarity 0.91)
- Mitigating Pervasive Modality Absence Through Multimodal Generalization and Refinement (AAAI 2025 · similarity 0.88)
- Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness (NeurIPS 2026 · similarity 0.88)
- Knowledge Bridger: Towards Training-Free Missing Modality Completion (CVPR 2025 · similarity 0.86)
- Multimodal Prompt Optimization: Why Not Leverage Multiple Modalities for MLLMs (ICLR 2026 · similarity 0.85)
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.
