计算机与数字工程2026，Vol.54Issue(3)：612-616,651,6.DOI:10.3969/j.issn.1672-9722.2026.03.005
非独立同分布数据流下的持续学习语义分割方法
A Semantic Segmentation Method for Continual Learning Under Non-independent and Identically Distributed Data Streams
摘要
Abstract
In order to alleviate the catastrophic forgetting phenomenon that occurs when the semantic segmentation model is in-crementally updated with knowledge,a continual learning semantic segmentation approach under non-independent homogeneously distributed data streams is proposed.Firstly,generative adversarial network generation and web crawling are used as data sources for the model,and old data are replayed during training to alleviate catastrophic forgetting.Secondly,to further optimise the knowl-edge recovery effect after replay,gating variables are introduced into the network,and a gating mechanism is built into the network to further improve model stability and plasticity.Experiments on the Pascal VOC 2012 dataset show that in the most complex incre-mental scenario 10-1,the mIoU of the initial class set and all classes improve by up to 2.5%and 2.2%compared to the baseline.关键词
非独立同分布数据流/持续学习/语义分割/灾难性遗忘Key words
non-independent and identically distributed data streams/continual learning/semantic segmentation/cata-strophic forgetting分类
信息技术与安全科学引用本文复制引用
李斌,于丽娅,杨静,李少波,袁坤..非独立同分布数据流下的持续学习语义分割方法[J].计算机与数字工程,2026,54(3):612-616,651,6.基金项目
国家自然科学基金项目(编号:62166005) （编号:62166005）
贵州省高层次留学人才项目(编号:(2021)09号) （编号:(2021）
贵州省自然科学基金项目(编号:黔科合基础-ZK[2022]一般130,黔科合支撑[2021]335,[2022]一般003) （编号:黔科合基础-ZK[2022]一般130,黔科合支撑[2021]335,[2022]一般003）
贵州大学人才引进项目(编号:贵大人基合字(2020)14号)资助. （编号:贵大人基合字(2020）
