Keyword spotting with phase-change memory matrix-vector multiplication accelerator

Daniele Ielmini (mail)

Intelligenza artificiale, robotica e computer vision

deep learning, emerging memory, hardware acceleration

15/05/2023

Python, Tensorflow, MATLAB

The goal of the project is to train and implement a convolutional neural network for keyword spotting on a custom integrated circuit, that features an embedded phase-change memory block for matrix-vector multiplication acceleration.