.png){kind=icon}
Return
Learnable multi-scale temporal convolutional network for multichannel time-series classification
.png){kind=icon}
PRE
AI
DOI：10.1016/j.engappai.2026.115572.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• A novel LMST-Net is proposed for gas-liquid two-phase flow pattern recognition. • MSDCB captures both microscopic bubble fluctuations and macroscopic flow evolution. • Learnable Feature Module aligns non-stationary signals via deformable convolution. • Temporal Attention suppresses noise in specific channels under low-SNR conditions. • An end-to-end framework handles raw signals without manual features.
Journal
IF:
8
Papers:
5.7K
Citations:
3.5W
Organization
Cited Papers
ShareSave
ShareSaveShareSaveApplication of Near-Infrared Spectroscopy to Measure the Water Content in Liquid Dielectrics.png)
Energies
OAAI
EnergiesIF0
OAAIShareSaveShareSave
View more