Description:
Fraunhofer Portugal has an IoT solution composed of hardware, firmware and software named Kallisto that aims to support this vision of integrating Machine-Learning-based algorithms into constrained devices.
Since it is unreasonable to support all existing frameworks, this thesis aims to test and evaluate existing frameworks to select the most promising ones and test them using the Kallisto micro-controller.
Outcome:
- Comparison between State-of-Art frameworks;
- Framework to support ML algorithms in Kallisto.
Author: Gonçalo Lemos
Type: MSc thesis
Partner: ISEP - Instituto Superior de Engenharia do Porto
Year: 2021