基于STM32CubeMX AI和NanoEdge AI的眼动信号分类效果对比研究
Comparative Study on the Classification Effect of Eye Movement Signals Based on STM32CubeMX AI and NanoEdge AI
禹鑫鹏 1贺庆 1王世昕1
作者信息
- 1. 北京信息科技大学仪器科学与光电工程学院,北京1 00096
- 折叠
摘要
为了实现在微处理器上运行眼动信号分类算法,精简嵌入式系统设计,提高系统效率,文章对比研究了STM32CubeMX AI和NanoEdge AI 2种可在微处理器上部署人工智能算法的技术手段.首先以眼电数据为基础,分别利用2种技术实现分类算法在微处理器上进行部署;然后在微处理器中运行分类算法,对眼电信号进行分类:最后对比分析2种分类方法的优缺点.实验结果表明,2种部署方式各有利弊,利用STM32CubeMXAI实现分类部署的方法首先需要在上位机中实现分类算法,有一定的执行难度,但可以更加有效地提高分类准确度;利用NanoEdge AI实现分类部署的方法可以避免上位机算法的调试,但无法实现针对不同信号进行具体设计.
Abstract
In order to achieve the implementation of eye movement signal classification algorithms on microprocessors,streamline embedded system design,and improve system efficiency,the article compares and studies two technical means that can deploy artificial intelligence algorithms on microprocessors:STM32CubeMX AI and NanoEdge AI.Firstly,based on the EEG data,machine learning,deep learning,and other methods are used to classify the EEG data in the upper computer;Then,two methods are used to implement the deployment and operation of classification algorithms on microprocessors;Finally,compare and analyze the advantages and disadvantages of the two classification methods.The experimental results show that both deployment methods have their own advantages and disadvantages.The method of using STM32CubeMX AI to achieve classification deployment requires the implementation of classification algorithms in the upper computer,which has certain execution difficulties but can more effectively improve classification accuracy;The method of implementing classification deployment using NanoEdge AI can avoid debugging the upper computer algorithm,but it cannot achieve high-precision classification.
关键词
眼电信号/嵌入式/STM32CubeMX/AI/NanoEdge/AI/人工智能Key words
EEG signal/embedded/STM32CubeMX AI/NanoEdge AI/artificial intelligence引用本文复制引用
出版年
2024
NETL
NSTL
万方数据