Fang, H.-S., Y. Xie, D. Shao, and C. Lu. “DIRV: Dense Interaction Region Voting for End-to-End Human-Object Interaction Detection”. Proceedings of the AAAI Conference on Artificial Intelligence, vol. 35, no. 2, May 2021, pp. 1291-9, doi:10.1609/aaai.v35i2.16217.