You are here

On-Device Training of Machine Learning Models on Microcontrollers With a Look at Federated Learning

TitleOn-Device Training of Machine Learning Models on Microcontrollers With a Look at Federated Learning
Publication TypeConference Paper
Year of Publication2021
AuthorsGrau, MMonfort, Pueyo-Centelles, R, Freitag, F
Conference NameProceedings of the Conference on Information Technology for Social Good
PublisherAssociation for Computing Machinery
Conference LocationNew York, NY, USA
ISBN Number9781450384780
Keywordsembedded systems, federated learning, keyword spotting, machine learning
AbstractRecent progress in machine learning frameworks makes it now possible to run an inference with sophisticated machine learning models on tiny microcontrollers. Model training, however, is typically done separately on powerful computers. There, the training process has abundant CPU and memory resources to process the stored datasets. In this work, we explore a different approach: training the model directly on the microcontroller. We implement this approach for a keyword spotting task. Then, we extend the training process using federated learning among microcontrollers. Our experiments with model training show an overall trend of decreasing loss with the increase of training epochs.
URLhttps://doi.org/10.1145/3462203.3475896
DOI10.1145/3462203.3475896