IP Library › Granted Patent US 12,657,957
Granted Patent B2
US 12,657,957 · App. 18/939,631 · Granted Jun 16, 2026

Data processing system and data processing method

Inventors: Tatsuya Okano (Isehara, JP); Teppei Oguni (Atsugi, JP); Kengo Akimoto (Isehara, JP)
Assignee: Semiconductor Energy Laboratory Co., Ltd.
G06V40/20G06V10/761G06V10/762G06V10/82G06V40/178
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,657,957
App. No.
18/939,631
Granted
Jun 16, 2026
Kind
B2
Abstract

A data processing system that can sense fatigue or the like using a neural network is provided. First, a reference image is obtained on the basis of first to n-th images (n is an integer greater than or equal to 2). Next, the first to n-th images and the reference image are input to an input layer of a neural network, first to n-th estimated ages and a reference estimated age are output from an output layer, and first to n-th data and reference data are output from an intermediate layer. After that, first to n-th coordinates are obtained in each of which an x-coordinate is a value corresponding to a difference between the reference estimated age and the first to n-th estimated ages and a y-coordinate is a value corresponding to the degree of similarity between the reference data and the first to n-th data. Next, a query image is input to the input layer, a query estimated age is output from the output layer, query data is output from the intermediate layer, and query coordinates are obtained using the output results. Whether a person of a face included in the query image feels fatigue or the like is determined on the basis of the first to n-th coordinates and the query coordinates.

Claims (44)

1 . A data processing system comprising:

an imaging portion;

a first processing portion;

a second processing portion;

a third processing portion; and

a fourth processing portion,

wherein the imaging portion is configured to obtain first to n-th images (n is an integer greater than or equal to 2) and a query image, each comprising an image of a user's face,

wherein the first processing portion is configured to obtain an average image on the basis of the first to n-th images,

wherein the second processing portion comprises a neural network comprising an input layer, an intermediate layer, and an output layer,

wherein the second processing portion is configured to output first to n-th estimated ages and first to n-th data comprising feature values of the first to n-th images in the case where the first to n-th images are input to the input layer of the neural network,

wherein the second processing portion is configured to output a reference estimated age and reference data comprising feature values of the average image in the case where the average image is input to the input layer of the neural network,

wherein the second processing portion is configured to output a query estimated age and query data comprising feature values of the query image in the case where the query image is input to the input layer of the neural network,

wherein the third processing portion is configured to obtain first to n-th coordinates using values of differences between the reference estimated age and each of the first to n-th estimated ages and values of degree of similarity between the reference data and each of the first to n-th data,

wherein the third processing portion is configured to obtain query coordinates using a value of a difference between the query estimated age and the reference estimated age and is a value of a degree of similarity between the query data and the reference data, and

wherein the fourth processing portion is configured to perform clustering on the basis of the first to n-th coordinates and determine whether the user feels stress on the basis of a result of the clustering and the query coordinates.

2 . The data processing system according to claim 1 , further comprising a memory portion,

wherein the memory portion is configured to store images obtained by the imaging portion and store data output from the first to fourth processing portions.

3 . A data processing method, comprising the steps of:

obtaining first to n-th images (n is an integer greater than or equal to 2) each comprising an image of a user's face;

obtaining an average image on the basis of the first to n-th images;

inputting the first to n-th images to an input layer of a neural network, thereby obtaining first to n-th estimated ages and first to n-th data comprising feature values of the first to n-th images from the neural network;

inputting the average image to the input layer of the neural network, thereby obtaining a reference estimated age and reference data comprising feature values of the average image from the neural network;

obtaining first to n-th coordinates using values of differences between the reference estimated age and each of the first to n-th estimated ages and values of degree of similarity between the reference data and each of the first to n-th data;

obtaining a query image comprising an image of the user's face;

inputting the query image to the input layer of the neural network, thereby obtaining a query estimated age and query data comprising feature values of the query image from the neural network;

obtaining query coordinates using a value of a difference between the query estimated age and the reference estimated age and a value of degree of similarity between the query data and the reference data; and

performing clustering on the basis of the first to n-th coordinates and determining whether the user feels stress on the basis of a result of the clustering and the query coordinates.

4 . An electronic device comprising:

an imaging portion;

a first processing portion;

a second processing portion;

a third processing portion; and

a fourth processing portion,

wherein the imaging portion is configured to obtain first to n+1-th images (n is an integer greater than or equal to 2) each comprising an image of a user's face,

wherein the first processing portion is configured to obtain an average image on the basis of the first to n-th images,

wherein the second processing portion comprises a neural network,

wherein the second processing portion is configured to output first to n-th estimated ages and first to n-th data comprising feature values of the first to n-th images in the case where the first to n-th images are input to the neural network,

wherein the second processing portion is configured to output a reference estimated age and reference data comprising feature values of the average image in the case where the average image is input to the neural network,

wherein the second processing portion is configured to output an n+1-th estimated age and n+1-th data comprising feature values of the n+1-th image in the case where the n+1-th image is input to the neural network,

wherein the third processing portion is configured to obtain first to n-th coordinates using values of differences between the reference estimated age and each of the first to n-th estimated ages and values of degree of similarity between the reference data and each of the first to n-th data,

wherein the third processing portion is configured to obtain n+1-th coordinates using a value of a difference between the n+1-th estimated age and the reference estimated age and a value of degree of similarity between the n+1-th data and the reference data, and

wherein the fourth processing portion is configured to perform clustering on the basis of the first to n-th coordinates and determine whether the user feels stress on the basis of a result of the clustering and the n+1-th coordinates.

5 . The electronic device according to claim 4 , further comprising a memory portion,

wherein the memory portion is configured to store images obtained by the imaging portion and store data output from the first to fourth processing portions.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Nov 7, 2024
From: OKANO, TATSUYA; OGUNI, TEPPEI; AKIMOTO, KENGO
To: SEMICONDUCTOR ENERGY LABORATORY CO., LTD.
Reel/Frame 069178/0140 →
Priority Claims (1)
JP 2019-221620 · Dec 6, 2019 · national
Continuity (2)
Continuation 17778623
Related Publication 20250061746A1 · Feb 20, 2025
References Cited (17)
US 10956716B2 · Martin et al. · 2021 [cited by applicant]
US 20170127992A1 · Takahashi · 2017 [cited by applicant]
US 20180330178A1 · el Kaliouby · 2018 [cited by examiner]
US 20190147156A1 · Burri · 2019 [cited by examiner]
US 20200219295A1 · el Kaliouby · 2020 [cited by examiner]
US 20200285842A1 · Wang · 2020 [cited by examiner]
US 20210049349A1 · Farokhi · 2021 [cited by examiner]
EP 3534296A · 2019 [cited by applicant]
JP 2007129312A · 2007 [cited by applicant]
JP 2010088756A · 2010 [cited by applicant]
JP 2011060038A · 2011 [cited by applicant]
JP 2014178229A · 2014 [cited by applicant]
JP 2016093313A · 2016 [cited by applicant]
JP 2017086524A · 2017 [cited by applicant]
JP 2019169139A · 2019 [cited by applicant]
International Search Report (Application No. PCT/IB2020/060947) Dated Feb. 22, 2021. [cited by applicant]
Written Opinion (Application No. PCT/IB2020/060947) Dated Feb. 22, 2021. [cited by applicant]