PAIRED NEURAL NETWORK FOR DIAGNOSING HEALTH STATE THROUGH VOICE

Japanese Patent JP2023038924

A

To improve health diagnosis by processing human voice.SOLUTION: A method includes: a step 610 of receiving a first audio signal corresponding to a first period; a step 620 of calculating a first feature vector from the first audio signal; a step 630 of processing the first feature vector to calculate a first voice embedded vector; a step 640 of receiving a second audio signal corresponding to a second period; a step 650 of calculating a second feature vector from the second audio signal; a step 660 of processing the second feature vector to calculate a second voice embedded vector; a step 670 of calculating element-wise difference between the first voice embedded vector and the second voice embedded vector; and a step 680 of using a mathematical model to process the element-wise difference, and thus calculating a variation value indicative of change in health state between the first period and the second period.SELECTED DRAWING: Figure 6

WO/2024/054937	OPTICAL COHERENCE TOMOGRAPHY SCANNING SYSTEM AND METHODS
WO/2017/184705	METHOD AND SYSTEM FOR MEASURING PERSPIRATION
JP6538249	Biometric information measuring device, biometric information display method, and program

SAMUEL KIM
KWON NAMHEE
NATHAN BLAYLOCK
HENRY J O'CONNELL
JEFFREY P ADAMS

JP2022139730A

March 17, 2023

September 02, 2022

Click for automatic bibliography generation  

CANARY SPEECH LLC

A61B5/00; G10L25/66; A61B5/16; G06N3/02; G10L15/00; G10L15/10; G16H50/20

Osamu Yamamoto
Toru Miyamae
Junichi Matsuo

Previous Patent: motor drive system

Next Patent: Beam forming device and beam control method