Research map / Graph learning / Graph algorithms and clustering
Ada-NETS: Face Clustering via Adaptive Neighbour Discovery in the Structure Space
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 Ada-NETS rather than STAR-FC | face clustering · MS-Celeb-1M, 584K unlabeled | BCubed F-score: 90.21 -> 91.40 | Helped |
| Use Ada-NETS rather than STAR-FC | face clustering · MS-Celeb-1M, 5.21M unlabeled | BCubed F-score: 81.47 -> 83.28 | Helped |
| Use Ada-NETS rather than STAR-FC | clustering · DeepFashion | Pairwise F-score: 37.07 -> 39.30 | Helped |
| Use Ada-NETS rather than STAR-FC | clustering · MSMT17 | Pairwise F-score: 58.80 -> 64.05 | Helped |
| Add adaptive neighbour discovery (ablation) | Ada-NETS graph building and clustering · MS-Celeb-1M ablation | Pairwise F-score: 77.17 -> 89.29 | Helped |
| Add structure space (ablation) | Ada-NETS graph building and clustering · MS-Celeb-1M ablation | Pairwise F-score: 77.17 -> 81.49 | Helped |
| Combine structure space with adaptive neighbour discovery (ablation) | Ada-NETS · MS-Celeb-1M ablation | Pairwise F-score: 89.29 -> 92.79 | Helped |
| Regress the optimal neighbour count directly rather than estimate Q values (ablation) | Ada-NETS adaptive filter · MS-Celeb-1M part_1 | Pairwise F-score: 86.10 -> 92.79 | Helped |
| Regress the optimal neighbour count rather than classify it (ablation) | Ada-NETS adaptive filter · MS-Celeb-1M part_1 | Pairwise F-score: 92.55 -> 92.79 | No clear effect |
| Replace GCN clustering with graph-cut (ablation) | Ada-NETS in the original feature space · MS-Celeb-1M part_1 | Pairwise F-score: 92.79 -> 72.95 | Hurt |
| Replace GCN clustering with graph-cut (ablation) | Ada-NETS in structure space · MS-Celeb-1M part_1 | Pairwise F-score: 92.79 -> 82.36 | Hurt |
| Use graph embeddings instead of original feature embeddings (ablation) | Ada-NETS · MS-Celeb-1M, 584K unlabeled | Pairwise F-score: 92.79 -> 93.74 | Helped |
Closest papers by meaning
- Learning to Cluster Faces via Confidence and Connectivity Estimation (CVPR 2020 · similarity 0.90)
- FC-DGCN: Deep Graph Convolutional Network for Face Clustering and Recognition (NeurIPS 2026 · similarity 0.89)
- Linkage Based Face Clustering via Graph Convolution Network (CVPR 2019 · similarity 0.89)
- Learning to Cluster Faces on an Affinity Graph (CVPR 2019 · similarity 0.88)
- Structure-Aware Face Clustering on a Large-Scale Graph With 107 Nodes (CVPR 2021 · similarity 0.88)
- Deep Contrastive Graph Learning with Clustering-Oriented Guidance (AAAI 2024 · similarity 0.82)
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.