CPC G06T 13/40 (2013.01) [G06F 3/017 (2013.01); G06T 17/00 (2013.01); G06V 20/52 (2022.01)] | 16 Claims |
1. A method comprising:
providing, by at least one processor, an emotional state model trained to predict an emotional state of an Artificial Intelligence (AI) character generated by an AI character model in a virtual environment;
continuously monitoring, by the at least one processor, a context of an interaction involving the AI character, wherein the continuously monitoring the context includes:
receiving, by the at least one processor, from a user, a message during a conversation of the AI character and the user; and
analyzing, by the at least one processor, a part of the message to determine the context, wherein the analysis of the part starts prior to receiving a next part of the message;
generating, by the at least one processor and based on the context, an input for the emotional state model;
providing, by the at least one processor, the input to the emotional state model;
obtaining, by the at least one processor, from the emotional state model, the emotional state of the AI character based on the input;
adjusting, by the at least one processor and based on the emotional state, parameters of the AI character model, thereby causing the AI character to act according to the emotional state;
receiving, by the at least one processor, the next part of the message;
analyzing, by the at least one processor, the next part of the message to determine a further context;
determining, by the at least one processor and based on the further context, a further input for the emotional state model;
providing, by the at least one processor, the further input to the emotional state model to obtain a further emotional state of the AI character; and
adjusting, by the at least one processor and based on the further emotional state, the parameters of the AI character model, thereby causing the AI character to act according to the further emotional state.
|