SemiSegECG
收藏资源简介:
SemiSegECG是一个为半监督ECG划线语义分割而建立的标准化基准。它由多个公共数据集组成,包括以前未被充分利用的资源。这个数据集旨在支持鲁棒和多样化的评估,并包含五个代表性的半监督分割算法,用于在两个不同的架构上进行评估:卷积网络和Transformer。此外,还提出了ECG特定的训练配置和增强策略,并引入了标准化的评估框架。这个基准将为半监督ECG划线方法的进步提供基础,并促进该领域进一步的研究。
SemiSegECG is a standardized benchmark established for semi-supervised ECG line semantic segmentation. It comprises multiple public datasets, including previously underutilized resources. This dataset aims to support robust and diverse evaluations, and incorporates five representative semi-supervised segmentation algorithms for evaluation across two distinct architectures: convolutional networks and Transformers. In addition, ECG-specific training configurations and augmentation strategies are proposed, and a standardized evaluation framework is introduced. This benchmark will lay a foundation for the advancement of semi-supervised ECG line segmentation approaches and promote further research in this field.




