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

Marc Monfort Grau, Roger Pueyo Centelles, Felix Freitag
2021 Proceedings of the Conference on Information Technology for Social Good  
Recent 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
more » ... end 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.
doi:10.1145/3462203.3475896 fatcat:5jasw4abk5ezppdg4cwpzzahzy