Systems and methods to facilitate bi-directional artificial intelligence communications

Number: 10635665
Published: 2020-04-28
Filed: 2017-12-21
Assignee: Disney Enterprises, Inc.
Inventors: Goslin; Michael P.
CPC: G06N5/022; G10L25/63; G10L15/22; G10L15/1815; G06F16/24522; G06F3/167
Verdict: Medium Notable software
Source: Google Patents · FreePatentsOnline

The keeper's note

AI conversational system for interactive characters.

Abstract

This disclosure presents systems and methods to facilitate artificial intelligence communications. One or more sensors may be configured to generate output signals conveying user behavior information and/or other information. The user behavior information may represent one or more communicative behaviors of a user in a real world. The one or more communicative behaviors may be interpreted based on communication structures of the one or more communicative behaviors to determine a meaning of a message conveyed by the one or more communicative behaviors. The communication structures may be categorized by structure type. The structure types may include one or more of a verbal type communication structure and/or a non-verbal type communication structure. Verbal type communication structure may refer to the conveyance of words via speech and/or non-speech communications. Non-verbal communication structure may refer to the conveyance of one or more of feelings, emotions, and/or impressions.

Background

FIELD(1) This disclosure relates to systems and methods to provide bi-directional artificial intelligence communications.BACKGROUND(2) Current virtual assistants (e.g., Siri, Watson, Alexa, etc.) may be designed to be intelligent, but not necessarily entertaining. These systems provide verbal answers/responses for questions the user poses to the assistant. These systems focus on text (or speech-to-text conversion) and/or speech processing to interact with a user.SUMMARY(3) One or more aspects of the disclosure relate to providing bi-directional artificial intelligence communications. Speech captures only a small subset of human communications. Much of what is communicated between humans is non-verbal and/or non-speech, such as facial expressions, physical gestures, tone/inflection of voice, and/or other nonverbal communications. These elements may be brought into the communication with intelligent assistants (or characters) to make the experience more realistic.(4) The communications received from a user may be interpreted to determine meanings of messages conveyed by those communications. The interpretation may be based on determining structure of communicative behaviors of the user in the real world and/or other information. The structure may include one or a combination of a verbal type communication structure and/or a non-verbal type communication structure. The verbal type communication structure may refer to the conveyance of words. Both speech and non-speech behaviors

Claims

1. A system configured to facilitate artificial intelligence (AI) communications, the system comprising: one or more physical processors configured by machine-readable instructions to: obtain output signals from one or more sensors, the output signals conveying user behavior information, the user behavior information representing one or more communicative behaviors of a user in a real world; determine individual communication structures of individual ones of the one or more communicative behaviors, the individual communication structures being an individual structure type that includes a verbal type communication structure or a non-verbal type communication structure; interpret the one or more communicative behaviors based on the individual communication structures to determine a meaning of a message conveyed by the one or more communicative behaviors; and obtain association information, the association information specifying associations between individual communicative behaviors and individual meanings of messages and associations between combinations of communicative behaviors and individual meanings of messages, the association information including: an association between a communicative behavior of the verbal type communication structure and a first meaning; and an association between a combination including a communicative behavior of the non-verbal type communication structure and the communicative behavior of the verbal type communication structure and a second meaning. 13. A method to facilitate artificial intelligence (AI) communications, the method comprising: obtaining output signals from one or more sensors, the output signals conveying user behavior information, the user behavior information representing one or more communicative behaviors of a user in a real world; determining individual communication structures of individual ones of the one or more communicative behaviors, the individual communication structures being an individual structure type that includes a verbal type communication structure or a non-verbal type communication structure; interpreting the one or more communicative behaviors based on the individual communication structures of the one or more communicative behaviors to determine a meaning of a message conveyed by the one or more communicative behaviors; and obtaining association information, the association information specifying associations between individual communicative behaviors and individual meanings of messages and associations between combinations of communicative behaviors and individual meanings of messages, the association information including: an association between a communicative behavior of the verbal type communication structure and a first meaning; and an association between a combination including a communicative behavior of the non-verbal type communication structure and the communicative behavior of the verbal type communication structure and a second meaning.