Skip to main content

Table 1 Performance comparison between baseline models and PUMPNET

From: PUMPNET: a deep learning approach to pump operation detection

Model mIOU Precision Recall Accuracy F1 score
Neural NILM rectangle 0.550 0.378 0.828 0.780 0.519
Single load extraction 0.568 0.403 0.738 0.806 0.521
PUMPNET 0.926 0.979 0.963 0.992 0.971