01314nas a2200205   4500000000100000008004100001260005900042653002100101653002300122653002100145653002100166100001400187700001600201700001800217245010400235856004400339300001400383520069300397020001801090       2021                            d  bAssociation for Computing MachineryaNew York, NY, USA10aembedded systems10afederated learning10akeyword spotting10amachine learning1 aMarc Grau1 aRoger Pueyo1 aFelix Freitag00aOn-Device Training of Machine Learning Models on Microcontrollers With a Look at Federated Learning  uhttps://doi.org/10.1145/3462203.3475896  a198–2033 aRecent 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.  a9781450384780