Reliable human orientation estimation (HOE) is critical for autonomous agents to understand human intention and perform human-robot interaction (HRI) tasks. Great progress has been made in HOE under full observation. However, the existing methods easily make a wrong prediction under partial observation and give it an unexpectedly high probability. To solve the above problems, this study first develops a method that estimates orientation from the visible joints of a target person so that it is able to handle partial observation. Subsequently, we introduce a confidence-aware orientation estimation method, enabling more accurate orientation estimation and reasonable confidence estimation under partial observation. The effectiveness of our method is validated on both public and custom-built datasets, and it showed great accuracy and reliability improvement in partial observation scenarios. In particular, we show in real experiments that our method can benefit the robustness and consistency of the robot person following (RPF) task.

翻译：可靠的人体朝向估计（HOE）对于自主智能体理解人类意图并执行人机交互（HRI）任务至关重要。目前在全观测条件下HOE已取得显著进展，但现有方法在部分观测下容易做出错误预测并赋予其异常高的置信度。为解决上述问题，本研究首先开发了一种从目标人物可见关节点估计朝向的方法，使其能够处理部分观测场景。随后，我们引入了一种置信度感知的朝向估计方法，能在部分观测下实现更精确的朝向估计与合理的置信度评估。通过在公开数据集和自建数据集上的验证，该方法在部分观测场景中展现出显著的精度与可靠性提升。特别地，实际实验证明该方法能够提升机器人跟随行人（RPF）任务的鲁棒性与一致性。