摘要
痛苦表情通常是身体出现状况或者不舒服时表现出来的一种类似警告的信息。将传统的7种表情分为痛苦和非痛苦两种,即将开心和中性表情归为非痛苦表情,伤心、惊讶、生气、恶心、害怕归为痛苦表情。基于几种卷积神经网络进行痛苦表情识别,最终得到一种准确率有91.8%的网络模型。
The pain expression is usually a warning-like message when something is wrong or uncomfortable.In this paper,seven traditional expressions are divided into pain and non-pain.Happy and neutral expressions are classified as nonpain expressions,and sad,surprised,angry,disgusted and afraid are classified as painful expressions.Finally,a network model with 91.8%accuracy is obtained based on several convolutional neural networks for painful expression recognition.
出处
《工业控制计算机》
2022年第11期109-110,共2页
Industrial Control Computer
基金
国家自然科学基金(61672199)。
