Research map / Learning theory and algorithms / Clustering
RAC-DMVC: Reliability-Aware Contrastive Deep Multi-View Clustering Under Multi-Source Noise
Shihao Dong, Yue Liu, Xiaotong Zhou, Yuhui Zheng, Huiying Xu, Xinzhong 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 |
|---|---|---|---|
| use reliability-aware noise contrastive loss instead of traditional contrastive loss (ablation) | RAC-DMVC · Caltech101 (50% noise) | ARI: 39.9 -> 73.9 | Improved |
| use reliability-aware noise contrastive loss instead of false-negative-aware contrastive loss (ablation) | RAC-DMVC · Caltech101 (50% noise) | ARI: 51.1 -> 73.9 | Improved |
| use reliability-aware noise contrastive loss instead of traditional contrastive loss (ablation) | RAC-DMVC · Scene15 (50% noise) | ACC: 40.8 -> 42.1 | Improved |
| add dual-attention imputation instead of direct imputation (ablation) | RAC-DMVC · NUS-WIDE (50% noise) | ACC: 34.4 -> 55.9 | Improved |
| add dual-attention imputation instead of k-nn imputation (ablation) | RAC-DMVC · Reuters (50% noise) | ACC: 57.6 -> 61.7 | Improved |
| add dual-attention imputation instead of prototype imputation (ablation) | RAC-DMVC · NUS-WIDE (50% noise) | ACC: 52.0 -> 55.9 | Improved |
| add noise contrastive learning module (ablation) | RAC-DMVC · 5 datasets (50% noise) | ACC, NMI, ARI | Improved |
| add cross-view reconstruction module (ablation) | RAC-DMVC · 5 datasets (50% noise) | ACC, NMI, ARI | Improved |
| add self-supervised cluster distillation module (ablation) | RAC-DMVC · 5 datasets (50% noise) | ACC, NMI, ARI | Improved |
| use RAC-DMVC instead of the best baseline at high noise | RAC-DMVC · LandUse21 and Scene15 (50% and 80% noise) | ACC: Scene15 80%: 32.4 vs SURE 36.0 | Mixed |
| reduce graph scaling factor sigma to 0.1 or below (ablation) | RAC-DMVC · 5 datasets (50% noise) | ACC, NMI | Improved |
| use smaller contrastive and distillation temperatures (ablation) | RAC-DMVC · Reuters, NUS-WIDE (50% noise) | ACC, NMI | Improved |
Closest papers by meaning
- Automatically Identify and Rectify: Robust Deep Contrastive Multi-view Clustering in Noisy Scenarios (ICML 2025 · similarity 0.92)
- Investigating and Mitigating the Side Effects of Noisy Views for Self-Supervised Clustering Algorithms in Practical Multi-View Scenarios (CVPR 2024 · similarity 0.90)
- EASEMVC:Efficient Dual Selection Mechanism for Deep Multi-View Clustering (CVPR 2025 · similarity 0.90)
- Self-supervised Trusted Contrastive Multi-view Clustering with Uncertainty Refined (AAAI 2025 · similarity 0.89)
- Robust Multi-view Clustering against Imperfect Information (NeurIPS 2026 · similarity 0.88)
- Medusa: A Multi-Scale High-order Contrastive Dual-Diffusion Approach for Multi-View Clustering (CVPR 2025 · similarity 0.88)
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.
