Browse Prior Art Database

Image Detection based evaluation of User Health Issues

IP.com Disclosure Number: IPCOM000250153D
Publication Date: 2017-Jun-06
Document File: 2 page(s) / 74K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a system that automatically detects an abnormal facial expression and/or posture within an image in relation to the surrounding context. The purpose is to help people with physical, mental, or emotional problems by initially detecting a possible problem through analysis of subtle abnormal facial expressions or postures, and then alerting the user when attention is needed from a health professional.

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

Image Detection based evaluation of User Health Issues

Disclosed is a system that automatically detects an abnormal facial expression and/or posture within an image in relation to the surrounding context. The purpose is to help people with physical, mental, or emotional problems by initially detecting a possible problem through analysis of subtle abnormal facial expressions or postures, and then alerting the user when attention is needed from a health professional.

In some situations, a person might not seem to have the appropriate facial expression or body posture for the mood or tone. For example, a person might look unhappy in a series of photographs capturing a celebration (e.g., wedding). In another example, a person might hold a strange pose with the head tilted to the side. If a person consistently shows these abnormal expressions and postures, then it might be cause to examine the physical and mental health of the individual. Further, a timely exam or evaluation can help the person find treatment or avoid an unpleasant series of event.

A method or system is needed to help people with physical, mental, or emotional problems by initially detecting a possible problem through analysis of subtle abnormal facial expressions or postures.

The novel contribution is a system that automatically detects an abnormal facial expression

and/or posture within an image in relation to the surrounding context. Based on the detection,

if the pattern reaches certain threshold (e.g., multiple instances over a period), then the system

recommends further medical diagnosis from selected health professionals.

The following detailed implementation describes two categories for facial expression and for

physical posture.

Category #1: Facial Expression

The system is integrated with the user's personal management software such as email,

calendar, instant messaging, and social networking website. The system can access the user's

profile data such as such as name, age, gender, etc. The system analyzes the content within the

user's images as well as the general context of images (e.g., user is with family on a vacation

trip, a class graduation photograph, etc.). The system analyzes the user’s facial expression with

the associated context and analyzes the sentiment based on the user's expression. The system

compares the user's sentiment and determines whether it matches the corresponding context.

The system builds a personalized data model to track a user's sentiment(s) and corresponding

context(s) along with the analysis results. The analysis results can include the positive outcome

or the negative outcome or no comparison outcome data. If the negative outcome pattern in

certain context exists many times and it exceeds the pre-defined threshold (e.g., 100 times)

then the system determines it is abnormal and intervention is...