<ul><li>Improving smart grid system management is crucial in the fight against climate change, allowing consumers to actively participate is a challenge for electricity suppliers.</li><li>Non-Intrusive Load Monitoring (NILM) aims to estimate power consumption of individual appliances using the main smart meter signal, which aggregates consumption making it hard to distinguish.</li><li>The newly introduced CamAL is a weakly supervised approach for appliance pattern localization, requiring only information on appliance presence to be trained.</li><li>CamAL combines deep-learning classifiers and an explainable classification method, outperforming existing weakly supervised baselines and demanding significantly fewer labels than fully supervised NILM approaches.</li></ul>

Few Labels are all you need: A Weakly Supervised Framework for Appliance Localization in Smart-Meter Series

Discover more