Research map / Vision-language and multimodal / Cross-modal learning
SeD-UD: An Influence-Driven and Hierarchically-Decoupled Information Bottleneck for Multimodal Intent Recognition
Qin Li, Wenbo Zhang, Limei Liu, Han Peng, Junfeng Yang, Guanying Xu
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 |
|---|---|---|---|
| replace fixed-dimension IB with input-adaptive IDAB (ablation) | SeD-UD · MIntRec | ACC (%): 71.47 -> 73.81 | Improved |
| use dynamic compression dimension instead of fixed average dimension (ablation) | IDAB with ranking selection · MIntRec | ACC (%): 69.99 -> 73.81 | Improved |
| use importance-ranking parameter selection instead of random (ablation) | IDAB with dynamic Dc · MIntRec | ACC (%): 71.91 -> 73.81 | Improved |
| add denoising on fused feature (ablation) | SeD-UD · MIntRec | ACC (%): 68.54 -> 71.69 | Improved |
| add unimodal de-redundancy (ablation) | SeD-UD · MIntRec | ACC (%): 68.54 -> 69.21 | Improved |
| combine de-redundancy on all three modalities with denoising (ablation) | SeD-UD · MELD-DA | ACC (%): 62.66 -> 63.72 | Improved |
| add text de-redundancy (ablation) | SeD-UD · MIntRec | ACC (%): 71.97 -> 73.81 | Improved |
| denoise after fusion (DR->MF->MD) instead of per-modality denoising (ablation) | SeD-UD · MIntRec | ACC (%): 71.24 -> 73.81 | Improved |
| perform de-redundancy before denoising (ablation) | SeD-UD · MIntRec | ACC (%): 70.63 -> 71.24 | Improved |
| use modality modulation-based fusion instead of linear mapping (ablation) | SeD-UD · MIntRec | ACC (%): 64.33 -> 73.81 | Improved |
| use modality modulation fusion instead of MAP (ablation) | SeD-UD · MIntRec | ACC (%): 72.93 -> 73.81 | Improved |
| set Dmin to 64 instead of 32 or 128 (ablation) | IDAB · MIntRec | ACC (%): 70.56 -> 73.81 | Improved |
Closest papers by meaning
- Adaptive Multimodal Fusion: Dynamic Attention Allocation for Intent Recognition (AAAI 2025 · similarity 0.87)
- MIntRec2.0: A Large-scale Benchmark Dataset for Multimodal Intent Recognition and Out-of-scope Detection in Conversations (ICLR 2024 · similarity 0.85)
- Contextual Augmented Global Contrast for Multimodal Intent Recognition (CVPR 2024 · similarity 0.84)
- Token-Level Contrastive Learning with Modality-Aware Prompting for Multimodal Intent Recognition (AAAI 2024 · similarity 0.83)
- Aligning Multimodal Representations through an Information Bottleneck (ICML 2025 · similarity 0.82)
- Decoupling Representation and Knowledge for Few-Shot Intent Classification and Slot Filling (AAAI 2024 · similarity 0.81)
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.
