Research map / Vision-language and multimodal / Cross-modal learning
CICA: Coupling Confidence-Aware Pretraining with Confidence-Informed Attention for Robust Multimodal Sentiment Analysis
Haoyu Jiang, Xiaoliang Chen, Duoqian Miao, Xiaolin Qin, Xianyong Li, Yajun Du
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 confidence-aware pretraining (CAP) (ablation) | CICA · MOSI | Corr: 0.791 -> 0.855 | Improved |
| add confidence-informed fusion (CIF) (ablation) | CICA · MOSI | Corr: 0.812 -> 0.855 | Improved |
| add confidence-aware pretraining (CAP) (ablation) | CICA · MOSEI | F1 (non-0): 83.19 -> 90.16 | Improved |
| add explicit CAP-CIF coupling (ablation) | CICA · MOSI | Corr: 0.831 -> 0.855 | Improved |
| add Mutual-information Contrastive Preservation (MCP) loss (ablation) | CICA · MOSI | Corr: 0.847 -> 0.855 | Improved |
| add intrinsic structure modulator Smod to attention (ablation) | CICA · MOSEI | Corr: 0.836 -> 0.856 | Improved |
| add MCP loss (ablation) | CICA · MOSEI | MAE: 0.496 -> 0.489 | Improved |
| add visual and audio modalities to text-only model (ablation) | CICA · MOSI | Corr: 0.857 (T only) vs 0.855 (T+V+A) | No clear effect |
| add audio modality to text and video (ablation) | CICA · MOSEI | MAE: 0.483 (T+A) vs 0.489 (T+V+A) | No clear effect |
| remove text modality (V+A only) (ablation) | CICA · MOSI | Corr: 0.855 -> 0.247 | Hurt |
| add Gaussian noise to audio on MOSEI (mild) | CICA · MOSEI | Corr: 0.856 -> 0.843 | Hurt |
| heavy video noise (0.8) on MOSI | CICA · MOSI | Corr: 0.855 -> 0.013 | Hurt |
Closest papers by meaning
- SemMSA: Latent Semantic-Aided Robust Multimodal Sentiment Analysis with Incomplete Data (NeurIPS 2026 · similarity 0.88)
- A Multi-Focus-Driven Multi-Branch Network for Robust Multimodal Sentiment Analysis (AAAI 2025 · similarity 0.88)
- A Unified Self-Distillation Framework for Multimodal Sentiment Analysis with Uncertain Missing Modalities (AAAI 2024 · similarity 0.88)
- Towards Robust Multimodal Sentiment Analysis with Incomplete Data (NeurIPS 2024 · similarity 0.87)
- Toward Robust Incomplete Multimodal Sentiment Analysis via Hierarchical Representation Learning (NeurIPS 2024 · similarity 0.87)
- DLF: Disentangled-Language-Focused Multimodal Sentiment Analysis (AAAI 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.
