arXiv · cs.AI · Jonathan Adam Rico, Nagarajan Raghavan, Senthilnath Jayavelu · 7 小时前模型评分31
JEPA-HITS：面向异构工业时间序列的跨域联合嵌入预训练
Cross-Domain Joint-Embedding Pretraining For Heterogeneous Industrial Time Series
导读
研究者提出跨域自监督预训练方法 JEPA-HITS，用联合嵌入预测架构（JEPA，Joint-Embedding Predictive Architecture）从电池、半导体化学机械抛光（CMP，Chemical Mechanical Planarization）、数控（CNC，Computer Numerical Control）加工和轴承等多类工业数据中训练单一编码器。
分类置信度
中 · 待人工复核
- 只有单一媒体来源，需要人工复核。
来源级别：媒体与个人
应用领域：制造与工业
事件类型：研究发布
证据摘录
- “We investigate cross-domain self-supervised pretraining using joint-embedding predictive architecture for heterogeneous industrial time series (JEPA-HITS), which learns one encoder from various domains such as battery, semiconductor chemical mechanical planarization (CMP), computer numerical control (CNC) machining, and bearing while preserving dataset-specific temporal scales.”
- “Comments: Submitted to International Conference on Acoustics, Speech, and Signal Processing (ICASSP) on Sept. 24, 2026”
来源：arXiv · cs.AI · arxiv.org