Browse Prior Art Database

Method and Apparatus for Customizable Composite Behavioral Recognition and Emotional States Mapping

IP.com Disclosure Number: IPCOM000246323D
Publication Date: 2016-May-30
Document File: 4 page(s) / 90K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed are an apparatus, system, and method to enable the automatic determination of a living subject’s emotional state based on the subject’s behavior.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 4

Method and Apparatus for Customizable Composite Behavioral Recognition and Emotional States Mapping

Studies on humans and animals show that behavior can reflect emotional states and physical needs . For example, if a dog's tail is horizontal to the ground, it usually means the dog is alert, but not aggressive.

Typically, humans perform the assessments of the subject's (either animal or human) behavior in order to determine its meaning. The assessment may also depend on the experience of the examiner . In addition, the scope of knowledge is usually general, collective knowledge of a species, not individualized information. For example, a behavior or action of an individual dog might have special meaning for that animal but be an uncommon behavior for other dogs .

A customizable, automatic method or system for behavior assessment is needed.

The novel contribution is an apparatus, system, and method to enable the automatic determination of a living subject's emotional state based on the subject's behaviors. The system collects and analyzes sensory and multi-media data (e.g., videos, pictures, audio, etc.).

The system categorizes the behavioral patterns into four categories: body movement, vocal expression, spatial movement, and facial expression. The system trains the visual and audio cognitive Application ProgrammingInterface (API) to recognize the behaviors of the living subject. It has a behavior-to-emotion mapping knowledge base for looking up the emotional states based on the recognized set of behaviors.

With this system, the user can automatically recognize the behaviors of a living subject through Audio and Visual Recognition cognitive APIs. The user can identify the emotional states based on the recognized behaviors and customize the system for the selected examined subject by providing feedback on the correctness of the assessment from any location at any time .

For illustration purposes, this disclosure categorizes the behaviors of living subjects (can be animal or human) into four groups:

Body movement (e.g., dog wagging its tail)

Vocal expression(e.g., dog barks)

Spatial movement (e.g., pacing around a room, jumping up)

Facial expression (e.g., mouth open, showing teeth, pupils dilated)

1


Page 02 of 4

The behavioral/emotional mapping knowledge base is capable of storing mappings of different species , such as human, dogs, and cats. The scenario described below is using dog as an example of a living subject. This can be applied to any living subjects.

Preparation/Training Phase

1. Add the knowledge of the behaviors of dogs into the knowledge base. For all the behavioral pattern categories, add the knowledge of the behaviors of dogs into the knowledge base. (...