IPM-FM: A Foundation Model with Consensus Feature Selection for Industrial Process Monitoring
The industrial process monitoring foundation model (IPM-FM) first learns general-purpose representations from unlabeled industrial process data through self-supervised pretraining, then adapts to specific monitoring tasks using a small amount of task-labeled data, and finally produces calibrated predictions through an...
Liang Cao, Wei-De Liu, Yan Qin et al.
· 0 citations