机电工程技术2024，Vol.53Issue(2)：20-23,4.DOI:10.3969/j.issn.1009-9492.2024.02.004
基于RepNet的自闭症健康监测方法研究
Study on the Health Monitoring of Autism Spectrum Disorder Based on RepNet
摘要
Abstract
Autism spectrum disorders(ASD)patients exhibit deviations in social interaction,language communication,and interests,despite displaying normal behavior in these aspects.With the increasing number of individuals diagnosed with autism,there is an urgent need for early autism screening to facilitate timely provision of specialized treatment.Currently,autism screening heavily relies on parents filling out questionnaires and healthcare professionals conducting manual observations and assessments,a process known for its low efficiency and time-consuming nature.The application of machine algorithms in the field of autism screening has not been widely adopted.To enhance screening efficiency and accuracy,the effectiveness of the machine vision algorithm RepNet in identifying repetitive movements in autism patients is explored;videos containing repetitive and non-repetitive actions are analyzed to evaluate the accuracy of RepNet;through the analysis of videos featuring repetitive movements in autism patients,the RepNet's accuracy and matching capabilities in screening for repetitive actions associated with autism are assessed.The results show that RepNet exhibits extremely high effectiveness and accuracy in detecting repetitive behaviors in individuals with autism.关键词
自闭症谱系障碍/人机交互/RepNet算法/机器视觉/重复性动作监测Key words
autism spectrum disorders(ASD)/human-computer interaction/RepNet algorithm/machine vision/repetitive actions detection分类
信息技术与安全科学引用本文复制引用
郭莹莹,何嫕琦,周俊耀,谢佳意,张晓宇,廖建源,吴羽庭,温晓红,张春良..基于RepNet的自闭症健康监测方法研究[J].机电工程技术,2024,53(2):20-23,4.基金项目
国家自然科学基金项目(52275097 （）
52205090) （）
广东省普通高校青年创新人才项目(2022KQNCX059 （）
2022KQNCX058) （）
广州市基础与应用基础研究基金(2023A04J1641) （2023A04J1641）
广州大学国家创新训练项目(202211078101) （202211078101）
