ITPUB
Aug 7, 2020 · Artificial Intelligence
How BC‑GNN Improves Temporal Action Proposals with Boundary‑Content Graph Modeling
The paper introduces Boundary Content Graph Neural Network (BC‑GNN), a graph‑based approach that jointly models boundary and content predictions to generate more accurate temporal action proposals and reliable confidence scores, achieving state‑of‑the‑art results on ActivityNet‑1.3 and THUMOS‑14.
BC-GNNECCV2020Graph Neural Network
0 likes · 12 min read
