US 12,333,717 B2
Infrared image sequence-based sleep quality evaluation system and method comprising performing evaluation by a classifier and counting the sleep quality evaluation results to determine the largest number to judge good or poor sleep quality
Shuqiang Wang, Guangdong (CN); Senrong You, Guangdong (CN); Guobao Wu, Guangdong (CN); Yiqian Lu, Guangdong (CN); Fen Miao, Guangdong (CN); and Chitang Zhang, Guangdong (CN)
Assigned to SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY, Guangdong (CN)
Appl. No. 17/786,840
Filed by SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY, Guangdong (CN)
PCT Filed Dec. 17, 2019, PCT No. PCT/CN2019/126038
§ 371(c)(1), (2) Date Jun. 17, 2022,
PCT Pub. No. WO2021/120007, PCT Pub. Date Jun. 24, 2021.
Prior Publication US 2023/0022206 A1, Jan. 26, 2023
Int. Cl. A61B 5/00 (2006.01); A61B 5/08 (2006.01); G06T 7/00 (2017.01); G06V 10/764 (2022.01); G06V 10/774 (2022.01); G06V 10/82 (2022.01)
CPC G06T 7/0012 (2013.01) [A61B 5/0082 (2013.01); A61B 5/08 (2013.01); A61B 5/4815 (2013.01); A61B 5/7267 (2013.01); G06V 10/764 (2022.01); G06V 10/7753 (2022.01); G06V 10/82 (2022.01); G06T 2207/10016 (2013.01); G06T 2207/10048 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30196 (2013.01); G06T 2207/30232 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A sleep quality evaluation method based on infrared image sequences, comprising:
obtaining a plurality of respiratory infrared image sequences to be evaluated, wherein each of the plurality of respiratory infrared image sequences to be evaluated comprises a plurality of respiratory infrared image frames to be evaluated;
performing sleep quality evaluation on each of the plurality of respiratory infrared image sequences to be evaluated by a classifier to obtain a-sleep quality evaluation results corresponding to the plurality of respiratory infrared image sequences to be evaluated;
counting different sleep quality evaluation results according to the sleep quality evaluation results respectively corresponding to the plurality of respiratory infrared image sequences to be evaluated, and determining one sleep quality evaluation result accounting for a largest number in the different sleep quality evaluation results as a sleep quality evaluation result of a user,
wherein before the performing the sleep quality evaluation on each of the plurality of respiratory infrared image sequences to be evaluated by the classifier, the sleep quality evaluation method further comprises:
training the classifier through a tensorized ternary generative adversarial network; and
wherein the tensorized ternary generative adversarial network comprises a generator, the classifier and a discriminator; and the training the classifier through the tensorized ternary generative adversarial network comprises:
inputting one-dimensional random noise and a target label into the generator, and obtaining a first respiratory infrared image sequence carrying the target label through a tensor decomposition-based deconvolutional layer of the generator;
inputting the first respiratory infrared image sequence into the discriminator, and obtaining a discrimination result of the first respiratory infrared image sequence by the discriminator through a tensor decomposition-based network layer and a full connection layer of the discriminator;
training the generator according to the discrimination result;
acquiring an unlabeled second respiratory infrared image sequence;
inputting the unlabeled second respiratory infrared image sequence into the classifier, and acquiring a third respiratory infrared image sequence through a second-order pooling block, the tensor decomposition-based network layer and the full connection layer of the classifier, wherein the third respiratory infrared image sequence refers to a labeled second respiratory infrared image sequence;
acquiring a labeled fourth respiratory infrared image sequence;
training the discriminator according to the first respiratory infrared image sequence, the third respiratory infrared image sequence and the labeled fourth respiratory infrared image sequence, and acquiring a discrimination result of the third respiratory infrared image sequence by the discriminator; and
training the classifier according to the first respiratory infrared image sequence, the discrimination result of the third respiratory infrared image sequence by the discriminator and the fourth respiratory infrared image sequence.