System for Low-Latency Animation of Talking Heads

A method of rendering a talking entity on a computing device having a client cache capable of storing audio/visual data associated with rendering the talking entity, the method causing a computing device to perform steps comprising:

• A. generating a talking entity response to a user utterance;
• B. retrieving cached sentences associated with bridging delays in a dialog and cached templates for use with dialogs;
• C. if at least one sentence or template relates to the talking entity response:
• 1. using the at least one sentence or template to render at least a part of the talking entity response;
• and
• 2. receiving a portion of the response not stored in a cache at the computing device to render a complete talking entity response;
• D. if none of the sentences or templates relates to the talking entity response:
• 1. receiving the talking entity response to be rendered;
• and
• E. rendering the talking entity response on the computing device.

A system for rendering a talking entity, the system having a processor, a machine-readable storage medium storing functional modules, and a cache storing data associated with rendering the talking entity, the computing device comprising:

• A. a module controlling the processor to generate a talking entity response to a user utterance;
• B. a module controlling the processor to retrieve cached sentences associated with bridging delays in a dialog and cached templates for use with dialogs;
• C. if at least one sentence or template relates to the talking entity response:
• 1. a module controlling the processor to use the at least one sentence or template to render at least a part of the talking entity response;
• and
• 2. a module controlling the processor to receive a portion of the response not stored in a cache at the computing device to render a complete talking entity response;
• D. if none of the sentences or templates relates to the talking entity response:
• 1. a module controlling the processor to receive the talking entity response to be rendered;
• and
• E. a module controlling the processor to render the talking entity response on the computing device.