Newborn pain expression recognition method and system based on deep 3D residual network

Abstract

The present invention discloses a newborn pain expression recognition method and system based on a deep 3D residual network. The method comprises: establishing a newborn expression video library containing pain expression category tags, and dividing samples in the newborn expression video library into a training set and a verification set; constructing a deep 3D residual network for newborn pain expression recognition, pre-training the network by using a publicized large-scale video database with category tags to obtain initial weight parameter values, and then performing fine-tune on the network by using the samples in the training set and in the verification set in the newborn expression video library to obtain a trained network model; and inputting a to-be-tested newborn expression video segment into the trained network model for expression classification recognition, and obtaining a pain expression recognition result. According to the technical scheme of the present invention, a deep 3D residual network is used to extract temporal and spatial dynamic features capable of reflecting time information from the video, and the change of the facial expression can be better characterized, so that the accuracy of the classification recognition can be improved.

Description

technical field [0001] The invention relates to the fields of facial expression recognition and machine learning, in particular to a method and system for recognizing newborn pain expressions based on a deep 3D residual network. Background technique [0002] Scientific research has proved that newborns have the ability to perceive pain. Neonatal pain mainly comes from painful operations, including plantar blood sampling, arteriovenous puncture, endotracheal intubation, subcutaneous and intramuscular injection, etc. Repeated or continuous pain stimulation has a series of short-term and long-term serious effects on the growth and development of newborns, which will lead to hazards such as mental retardation, central nervous system damage, and emotional disorders in newborns. Pain assessment is an important part of pain control, so correctly assessing pain and taking corresponding analgesic measures in time to relieve neonatal pain has important clinical application value, whi...

AI Technical Summary

Problems solved by technology

[0007] Purpose of the invention: Aiming at the problems of the prior art, the purpose of the present invention is to provide a newborn pain expression recognition method and system based on a deep 3D residual network. The residual unit structure is applied to a deep convolutional neural network, which can effectively alleviate the network pain. The gradient disappearance problem in backpropagation during model training solves the problem of difficult training and performance degradation of deep networks. At the same time, the combined operation of 2D convolution and 1D convolution is used to realize 3D convolution operation. Compared with the same depth The 2D convolutional neural network only adds a certain number of 1D convolutions, and does not produce excessive growth in the number of parameters, running time, etc.

Images