南京邮电大学学报（自然科学版）2025，Vol.45Issue(5)：66-73,8.DOI:10.14132/j.cnki.1673-5439.2025.05.008
基于多尺度特征融合和对比学习的小样本图像分割方法
Few-shot image segmentation based on multi-scale feature fusion and contrastive learning
摘要
Abstract
In response to the scarcity of annotated medical image data and the imitations of existing mod-els in segmenting multi-scale target images,this paper proposes a few-shot medical image segmentation method based on multi-scale feature fusion and contrastive learning.First,a sequential concatenation-based multi-scale skip connection method is introduced to replace traditional skip connections,enabling effective fusion of multi-scale feature maps from the encoder and their transmission to the corresponding decoder.Second,considering the dual-branch structure of the model,a contrastive learning module based on multi-scale features is proposed,and a loss function is designed to enhance the model's dis-criminative ability at the pixel level.Experiments show that our method achieves cross-domain data seg-mentation for medical images,mitigates performance degradation due to dataset scarcity,and improves the segmentation accuracy and generalization for different-scale targets,outperforming current main-stream few-shot medical image segmentation methods.关键词
深度学习/医学图像分割/多尺度特征融合/对比学习/小样本学习Key words
deep learning/medical image segmentation/multi-scale feature fusion/contrastive learn-ing/few-shot learning分类
信息技术与安全科学引用本文复制引用
胡晓飞,吴佳芸,邹贵春,武灵芝..基于多尺度特征融合和对比学习的小样本图像分割方法[J].南京邮电大学学报（自然科学版）,2025,45(5):66-73,8.基金项目
国家自然科学基金(61771251)和江苏省高校实验室研究课题(GS2022YB36)资助项目 （61771251）
