Research map / Graph learning / Graph algorithms and clustering
Linkage Based Face Clustering via Graph Convolution Network
Check 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 |
|---|---|---|---|
| replace heuristic clustering with context-based GCN link prediction | Face clustering · IJB-B-512 | F-measure: 0.802 -> 0.833 | Helped |
| replace heuristic clustering with context-based GCN link prediction | Face clustering · IJB-B-1024 | F-measure: 0.805 -> 0.833 | Helped |
| replace heuristic clustering with context-based GCN link prediction | Face clustering · IJB-B-1845 | F-measure: 0.800 -> 0.814 | Helped |
| use attention aggregation instead of mean aggregation (ablation) | GCN face clustering · IJB-B-512 | F-measure: 0.821 -> 0.833 | Helped |
| use attention aggregation instead of weighted aggregation (ablation) | GCN face clustering · IJB-B-1845 | F-measure: 0.801 -> 0.814 | Helped |
| increase the 1-hop neighbor count from k_1=10 to k_1=160 (k_2=5) (ablation) | GCN face clustering · IJB-B-1845+1M | F-measure: 0.634 -> 0.720 | Helped |
| increase the 1-hop neighbor count from k_1=10 to k_1=160 (k_2=5) (ablation) | GCN face clustering · IJB-B-1845+1M | runtime: 00:19:50 -> 02:45:36 | Hurt |
| use more 1-hop candidates (ablation) | GCN face clustering · IJB-B-512 | F-measure | Helped |
| use more 2-hop context neighbors (ablation) | GCN face clustering · IJB-B-512 | F-measure | Helped |
| increase the number of hops beyond h=2 (ablation) | GCN face clustering · IJB-B-512 | F-measure | No clear effect |
| vary the linked-neighbor count u (ablation) | GCN face clustering · IJB-B-512 | F-measure | No clear effect |
| backpropagate supervision only for 1-hop neighbors rather than all neighbors (ablation) | GCN link prediction · Face clustering | accuracy | Helped |
Closest papers by meaning
- Learning to Cluster Faces via Confidence and Connectivity Estimation (CVPR 2020 · similarity 0.93)
- Structure-Aware Face Clustering on a Large-Scale Graph With 107 Nodes (CVPR 2021 · similarity 0.92)
- Learning to Cluster Faces on an Affinity Graph (CVPR 2019 · similarity 0.92)
- FC-DGCN: Deep Graph Convolutional Network for Face Clustering and Recognition (NeurIPS 2026 · similarity 0.91)
- Density-Aware Feature Embedding for Face Clustering (CVPR 2020 · similarity 0.89)
- Ada-NETS: Face Clustering via Adaptive Neighbour Discovery in the Structure Space (ICLR 2022 · similarity 0.89)
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.