Research map / Learning theory and algorithms / Clustering
Multi-VAE: Learning Disentangled View-Common and View-Peculiar Visual Representations for Multi-View Clustering
Jie Xu, Yazhou Ren, Huayi Tang, Xiaorong Pu, Xiaofeng Zhu, Ming Zeng, Lifang He
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 controlled capacity on the view-common variable (Cc) to the vanilla VAE variant (ablation) | Multi-VAE · Multi-COIL-10 | ACC: 0.872 -> 0.902 | Improved |
| add controlled capacity on both common and peculiar variables (Cc + Cz) instead of Cc only (ablation) | Multi-VAE · Multi-COIL-10 | ACC: 0.902 -> 0.993 | Improved |
| add controlled capacity on both variables instead of Cz only (ablation) | Multi-VAE · Multi-COIL-10 | ACC: 0.704 -> 0.993 | Improved |
| replace vanilla VAE with beta-VAE (higher KL weight, no capacity) (ablation) | Multi-VAE · Multi-COIL-10 | ACC: 0.872 -> 0.683 | Hurt |
| add controlled capacity only on view-peculiar variables (Cz) versus vanilla VAE (ablation) | Multi-VAE · Multi-COIL-10 | ACC: 0.872 -> 0.704 | Hurt |
| add controlled capacity (Cc + Cz) to beta-VAE (ablation) | Multi-VAE · Multi-COIL-10 | ACC: 0.683 -> 0.993 | Improved |
| concatenate view-peculiar representations with view-common (Multi-VAE-CZ with K-means) instead of using c alone | Multi-VAE · Multi-MNIST | ACC: 0.989 -> 0.999 | Improved |
| concatenate view-peculiar representations with view-common instead of c alone | Multi-VAE · Multi-Fashion | ACC: 0.816 -> 0.907 | Improved |
| use multi-view Multi-VAE-C instead of single-view beta-VAE for clustering | Multi-VAE-C · Multi-COIL-10 | ACC: 0.598 -> 0.900 | Improved |
Closest papers by meaning
- Learning Disentangled Representations for Generalized Multi-View Clustering (TPAMI 2026 · similarity 0.91)
- Deep Variational Incomplete Multi-View Clustering: Exploring Shared Clustering Structures (AAAI 2024 · similarity 0.89)
- EASEMVC:Efficient Dual Selection Mechanism for Deep Multi-View Clustering (CVPR 2025 · similarity 0.88)
- Deep Incomplete Multi-view Learning via Cyclic Permutation of VAEs (ICLR 2025 · similarity 0.88)
- Differentiable Information Bottleneck for Deterministic Multi-view Clustering (CVPR 2024 · similarity 0.87)
- Enhanced then Progressive Fusion with View Graph for Multi-View Clustering (CVPR 2025 · similarity 0.87)
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.
