Digital character with dynamic interactive behavior
A virtual experience system provides dynamically interactive virtual characters in an environment. The system may be built in a modular fashion, enabling new behaviors and functionality to be added without significant alterations to the existing parts of the system. For example, behavior may be modeled using a model with multiple layers. Each layer determines responses to external stimuli based on one or more factors, with the behavior of a character being determined based on a combination of the layers of the model.
1. A method of providing dynamic responses of characters to stimuli, the method comprising:
identifying a stimulus for a character;
obtaining contextual information associated with the character, wherein obtaining the contextual information associated with the character includes obtaining information about a current environment of the character, the information about the current environment of the character comprising an identity of an additional character in the environment and a relationship between the character and the additional character;
determining a response type of the character to the stimulus;
selecting a set of behavioral assets by applying, as input to a plurality of layers, input data derived from the response type and the contextual information;
determining a response of the character to the stimulus by combining at least some of the behavioral assets; and
providing a visualization of the response of the character to the stimulus for display to a user.
2. The method of claim 1 , wherein obtaining contextual information associated with the character comprises obtaining a character model including attributes of the character.
3. The method of claim 2 , wherein the attributes of the character includes a current energy level, and the behavioral assets in the set are selected based on the current energy level.
4. The method of claim 1 , wherein the stimulus includes at least one of: a message, a line of dialogue, a virtual event, a real-world event, an action of another character, an action of a real person, music, a sound effect, or an object perceived by the character.
5. The method of claim 1 , wherein the plurality of layers comprises at least one of: an idle layer, a breathing layer, a weight shift layer, a posture layer, a gestures layer, a performances layer, a look-ats layer, an interrupts layer, or a refusing layer.
6. The method of claim 1 , wherein determining the response comprises:
selecting a subset of the set of behavioral assets to trigger in response to the stimulus; and
blending the subset of assets.
7. The method of claim 6 , wherein each set of behavioral assets is associated with a corresponding probability of the behavioral asset triggering, and the subset is selected using a random number generator and the corresponding probabilities.
8. The method of claim 7 , wherein the probability is based on at least one of: a suitability of the behavioral asset for the character or a degree to which the character is experiencing a corresponding emotion.
9. The method of claim 1 , wherein the stimulus is music and the response is the character automatically performing one or more dance steps in time with the music.
10. The method of claim 1 , wherein the response includes a gesture and an amplitude of the gesture or action is modified by a degree of extroversion of the character.
11. The method of claim 1 , wherein the response includes a gesture and an amplitude of the gesture is modified by a current mood of the character.
12. The method of claim 1 , wherein the response includes a gesture and an amplitude of the gesture is modified by an energy level of the character.
13. The method of claim 1 , wherein the character is a digital being, a virtual being, a synthetic being, or a robotic being.
14. A method of providing dynamic responses of characters to stimuli, the method comprising:
identifying a stimulus for a character;
obtaining contextual information associated with the character;
determining a response type of the character to the stimulus;
selecting a set of behavioral assets by applying, as input to a plurality of layers, input data derived from the response type and the contextual information, wherein the plurality of layers comprises at least one of: an idle layer, a breathing layer, a weight shift layer, a posture layer, a gestures layer, a performances layer, a look-ats layer, an interrupts layer, or a refusing layer, and wherein the input data for at least one layer of the plurality of layers comprises demographic information for the character, and the at least one layer outputs one or more assets tagged as suitable for use with the character in view of the demographic information;
determining a response of the character to the stimulus by combining at least some of the behavioral assets; and
providing a visualization of the response of the character to the stimulus for display to a user.
15. The method of claim 14 , wherein obtaining the contextual information associated with the character includes obtaining information about a current environment of the character, the information about the current environment of the character comprising an identity of an additional character in the environment and a relationship between the character and the additional character.
16. A method of providing dynamic responses of characters to stimuli, the method comprising:
identifying a stimulus for a character;
obtaining contextual information associated with the character;
determining a response type of the character to the stimulus;
selecting a set of behavioral assets by applying, as input to a plurality of layers, input data derived from the response type and the contextual information, wherein a first asset is included in the set responsive to a value of a first attribute of the character exceeding a threshold;
determining a response of the character to the stimulus by combining at least some of the behavioral assets; and
providing a visualization of the response of the character to the stimulus for display to a user.
17. The method of claim 16 , wherein the threshold is adjusted by a random variation prior to comparing the value of the first attribute to the threshold.
18. The method of claim 16 , wherein obtaining the contextual information associated with the character includes obtaining information about a current environment of the character, the information about the current environment of the character comprising an identity of an additional character in the environment and a relationship between the character and the additional character.
19. A method of providing dynamic responses of characters to stimuli, the method comprising:
identifying a stimulus for a character;
obtaining contextual information associated with the character;
determining a response type of the character to the stimulus;
selecting a set of behavioral assets by applying, as input to a plurality of layers, input data derived from the response type and the contextual information;
determining a response of the character to the stimulus by combining at least some of the behavioral assets, wherein the response is determined using a natural language processing model that relates words to quick gestures, phrases to longer gestures, and sentences to pose changes; and
providing a visualization of the response of the character to the stimulus for display to a user.
20. The method of claim 19 , wherein obtaining the contextual information associated with the character includes obtaining information about a current environment of the character, the information about the current environment of the character comprising an identity of an additional character in the environment and a relationship between the character and the additional character.