logo
CanyamAI summaries for academic research
language
MobileMobile
  • Phone FeatureWeChat Mini Program
EnglishEnglish
noticeNotice
BlogBlog
delete
arrow
Return

DiffTCL: Diffusion-based augmentation for time series contrastive learning

delete2026-09-02
delete0
PRE
AI
Z
Zhongyang Yu
Z
Zhihai Wang
J
Jidong Yuan *
H
Huiting Pei
Y
Yuxuan Fan
M
Muhang Li
DOI：10.1016/j.patcog.2026.114833delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• DiffTCL is a diffusion framework for time-series contrastive learning. • DiffAu generates semantically consistent views while preserving global statistics. • DiffSM mines positive views and hard negatives to mitigate false negatives.
Keywords:
Time series
Diffusion model
Contrastive learning

Journal

Pattern Recognition cover
Pattern Recognition
IF:
7.6
Papers:
1.3W
Citations:
4.5W

Organization

No organization information available