Abstract
In order to enable a social agent to behave in a believable and realistic way, it needs a wide range of information in the form of both low-level value-based data as well as high-level semantic knowledge. In this work we propose a system that puts a virtual reality layer between the real world and an agent\u0027s knowledge representation. This mirror world allows the agent to use its abstract representation of the environment and inferred events as an additional source of knowledge when reasoning about the real world. Additionally, users and developers can use the mirror world, with its visualized data and highlighting of the agent\u0027s reasoning, for further understanding of the agent\u0027s behavior, debugging and testing, or the simulation of additional sensor input.
Users
Please
log in to take part in the discussion (add own reviews or comments).