Inference speed and quantisation of neural networks with TensorFlow Lite for Microcontrollers framework | IEEE Conference Publication | IEEE Xplore