Research map / Vision-language and multimodal / Cross-modal learning
Unsupervised Semantic Discovery via Global and Local Semantic Alignment in Multimodal Clustering
Zhengzhong Zhu, Pei Zhou, Weihong Du, Shiquan Min, Jiangping Zhu
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 global semantic alignment (GSA) (ablation) | GLAD · IEMOCAP-DA | Avg: 18.29 -> 32.17 | Improved |
| add global semantic alignment (GSA) (ablation) | GLAD · MIntRec | Avg: 23.90 -> 39.14 | Improved |
| add local semantic alignment (LSA) (ablation) | GLAD · MIntRec | Avg: 36.04 -> 39.14 | Improved |
| add local semantic alignment (LSA) (ablation) | GLAD · IEMOCAP-DA | Avg: 27.58 -> 32.17 | Improved |
| use cluster-based weights s_ij in LSA instead of plain same-sample alignment (ablation) | GLAD · MIntRec | Avg: 36.44 -> 39.14 | Improved |
| use cluster-based weights s_ij in LSA (ablation) | GLAD · MELD-DA | Avg: 28.89 -> 30.99 | Improved |
| use optimal transport instead of cosine similarity for assignments (ablation) | GLAD GSA · MIntRec | Avg: 35.80 -> 39.14 | Improved |
| use optimal transport instead of cosine similarity for assignments (ablation) | GLAD GSA · MELD-DA | Avg: 28.70 -> 30.99 | Improved |
| add OT component versus none (ablation) | GLAD GSA · MIntRec | Avg: 34.53 -> 39.14 | Improved |
| GLAD vs UMC | UMC · IEMOCAP-DA | Avg: 27.71 -> 32.17 | Improved |
| GLAD vs UMC | UMC · MIntRec | ACC: 43.73 -> 45.97 | Improved |
| predict clusters from label vector argmax instead of k-means on features (ablation) | GLAD · IEMOCAP-DA | clustering performance | Improved |
Closest papers by meaning
- On the Value of Cross-Modal Misalignment in Multimodal Representation Learning (NeurIPS 2025 · similarity 0.86)
- Beyond Global Similarity: Multi-Conditional Retrieval for Fine-Grained Cross-Modal Understanding (CVPR 2026 · similarity 0.86)
- Incorporating Dense Knowledge Alignment into Unified Multimodal Representation Models (CVPR 2025 · similarity 0.86)
- Topology-Aware Representation Alignment for Semi-Supervised Vision-Language Learning (NeurIPS 2026 · similarity 0.85)
- SAE-V: Interpreting Multimodal Models for Enhanced Alignment (ICML 2025 · similarity 0.85)
- Low-Rank Similarity Mining for Multimodal Dataset Distillation (ICML 2024 · 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.
