Intelligent Robot Which Can Recognize Occurrence of Occlusion

在本研究中，为了在经常发生视觉遮挡的环境中实现智能人机交互，我们开发了遮挡鲁棒跟踪方法、准确估计周围环境的方法以及定位被其他物体遮挡的声源的方法。此外，机器人行为印象评价的实验结果表明，采用上述技术开发的机器人能给人留下更熟悉、更愉快的印象。

In this study, in order to achieve intelligent human-robot interactions in environments where visual occlusion will often occurs, we have developed the occlusion-robust tracking method, the method for estimating the surrounding environments accurately, and the method for locating the sound sources which are occluded by other objects. Furthermore, experimental results of impression evaluation of robot's behaviors show that the developed robot which uses above techniques can give more familiar and pleasant impressions to humans.