Zobrazeno 1 - 2
of 2
pro vyhledávání: '"Hoshino, Hana"'
Multi-modal trajectory forecasting methods commonly evaluate using single-agent metrics (marginal metrics), such as minimum Average Displacement Error (ADE) and Final Displacement Error (FDE), which fail to capture joint performance of multiple inter
Externí odkaz:
http://arxiv.org/abs/2305.06292
Inverse Reinforcement Learning (IRL) is attractive in scenarios where reward engineering can be tedious. However, prior IRL algorithms use on-policy transitions, which require intensive sampling from the current policy for stable and optimal performa
Externí odkaz:
http://arxiv.org/abs/2109.04307