Implementation of the algorithm described in "Fully Distributed Deep Learning Inference on Resource-Constrained Edge Devices"