Research map / Label-efficient and robust learning / Recognition and long-tail
Reliable Neighborhood-Aware Multi-View Outlier Detection
Huijie Ma, Haoyuan Xin, Lei Meng, Guanzhou Ke, Yongyong Chen, Guoqing Chao
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 alignment loss, sample reliability and consistency reliability to reconstruction-only model (ablation) | RNAMOD · Caltech101 | AUC: 0.791 -> 0.891 | Improved |
| add leave-one-out directional consensus (ablation) | RNAMOD (SR+CR+AL) · six datasets | AUC: Caltech101 0.891 -> 0.884, NUSWIDEOBJ 0.884 -> 0.892 | Mixed |
| add view weighting (ablation) | RNAMOD (SR+CR+AL+LOO) · Caltech101 | AUC: 0.884 -> 0.898 | Improved |
| add sample reliability (SR) (ablation) | RNAMOD without SR · Caltech101 | AUC: 0.872 -> 0.898 | Improved |
| add consistency reliability (CR) (ablation) | RNAMOD without CR · Caltech101 | AUC: 0.879 -> 0.898 | Improved |
| build reliable neighborhoods instead of conventional kNN (ablation) | RNAMOD · Synthetic dataset | neighbor connections involving outliers | Improved |
| set alpha to 0 (remove sample reliability) or 1 (ablation) | RNAMOD · six datasets | AUC | Hurt |
| vary neighborhood size k and loss weight lambda (ablation) | RNAMOD · six datasets | AUC | Mixed |
| use RNAMOD instead of RCPMOD on 100Leaves with higher class-outlier ratio | RNAMOD · 100Leaves config 2 | AUC: 0.837 -> 0.767 | Hurt |
Closest papers by meaning
- Neighborhood Consensus Networks for Unsupervised Multi-view Outlier Detection (AAAI 2021 · similarity 0.91)
- Extremely Simple Multimodal Outlier Synthesis for Out-of-Distribution Detection and Segmentation (NeurIPS 2025 · similarity 0.85)
- Diffusion-based Semantic Outlier Generation via Nuisance Awareness for Out-of-Distribution Detection (AAAI 2025 · similarity 0.85)
- FlexUOD: The Answer to Real-world Unsupervised Image Outlier Detection (CVPR 2025 · similarity 0.84)
- ALTBI: Constructing Improved Outlier Detection Models via Optimization of Inlier-Memorization Effect (AAAI 2025 · similarity 0.84)
- DOS: Diverse Outlier Sampling for Out-of-Distribution Detection (ICLR 2024 · similarity 0.84)
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.
