AutoFlow is a tool that helps developers to implement machine learning (ML) faster and easier on embedded devices. The whole workflow of a data scientist should be covered. Starting from building the ML model to the selection of the target platform to the optimization and implementation of the model on the target platform. To realize these functions, AutoFlow was divided into two parts. One part represents the automatic generation of neural networks. This is realized by using Automated machine learning (AutoML). For a given data set, different neural networks are automatically trained, from which the one that achieves the highest accuracy is stored. In the other part of AutoFlow, neural networks can be reduced in size using pruning and quantization. Also, the target platform on which the model is to be executed later can be selected. Accordingly, the model is converted and the necessary files are generated for execution on the target platform. AutoFlow is an open-source tool and can be downloaded from GitHub.
The Tool is available at: https://github.com/Hahn-Schickard/AUTOflow