Research map / Video / Action recognition
Boundary Content Graph Neural Network for Temporal Action Proposal Generation
Yueran Bai, Yingying Wang, Yunhai Tong, Yang Yang, Qiyue Liu, Junhui Liu
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 |
|---|---|---|---|
| replace GCN with BC-GNN graph reasoning (undirected, no edge update) (ablation) | BC-GNN · ActivityNet-1.3 val proposals | AR@100: 75.57 -> 76.18 | Improved |
| replace GCN with BC-GNN graph reasoning (undirected, no edge update) (ablation) | BC-GNN · ActivityNet-1.3 val proposals | AUC: 66.88 -> 67.36 | Improved |
| add edge updating to undirected graph (ablation) | BC-GNN · ActivityNet-1.3 val proposals | AR@100: 76.18 -> 76.40 | Improved |
| add edge updating to undirected graph (ablation) | BC-GNN · ActivityNet-1.3 val proposals | AUC: 67.36 -> 67.79 | Improved |
| add directed edges without edge updating (ablation) | BC-GNN · ActivityNet-1.3 val proposals | AR@100: 76.18 -> 76.15 | No clear effect |
| add directed edges without edge updating (ablation) | BC-GNN · ActivityNet-1.3 val proposals | AUC: 67.36 -> 67.53 | Improved |
| add directed edges with edge updating (ablation) | BC-GNN · ActivityNet-1.3 val proposals | AR@100: 76.40 -> 76.73 | Improved |
| add directed edges with edge updating (ablation) | BC-GNN · ActivityNet-1.3 val proposals | AUC: 67.79 -> 68.05 | Improved |
| add both directed edges and edge updating (ablation) | BC-GNN · ActivityNet-1.3 val proposals | AR@100: 76.18 -> 76.73 | Improved |
| use BC-GNN proposals instead of BMN proposals for detection | two-stage detector · ActivityNet-1.3 val detection | average mAP: 33.85 -> 34.26 | Improved |
| use BC-GNN proposals instead of BMN proposals for detection (UNet) | two-stage detector · THUMOS-14 test detection, tIoU 0.5 | mAP: 38.8 -> 40.4 | Improved |
| Soft-NMS instead of NMS for BC-GNN | BC-GNN · THUMOS-14 test proposals (2Stream) | AR@100: 50.35 -> 49.60 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- CTAP — CTAP: Complementary Temporal Action Proposal Generation (ECCV 2018) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- BSN: Boundary Sensitive Network for Temporal Action Proposal Generation (ECCV 2018 · similarity 0.92)
- BMN: Boundary-Matching Network for Temporal Action Proposal Generation (ICCV 2019 · similarity 0.91)
- Fast Learning of Temporal Action Proposal via Dense Boundary Generator (AAAI 2020 · similarity 0.89)
- BSN++: Complementary Boundary Regressor with Scale-Balanced Relation Modeling for Temporal Action Proposal Generation (AAAI 2021 · similarity 0.89)
- Temporal Context Aggregation Network for Temporal Action Proposal Refinement (CVPR 2021 · similarity 0.87)
- DCAN: Improving Temporal Action Detection via Dual Context Aggregation (AAAI 2022 · 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.
