Result Details
Hardware-Aware Evolutionary Approaches to Deep Neural Networks
Mrázek Vojtěch, Ing., Ph.D., DCSY (FIT)
Piňos Michal, Ing., DCSY (FIT)
This chapter gives an overview of evolutionary algorithm (EA) based methods applied to the design of efficient implementations of deep neural networks (DNN). We introduce various acceleration hardware platforms for DNNs developed especially for energy-efficient computing in edge devices. In addition to evolutionary optimization of their particular components or settings, we will describe neural architecture search (NAS) methods adopted to directly design highly optimized DNN architectures for a given hardware platform. Techniques that co-optimize hardware platforms and neural network architecture to maximize the accuracy-energy trade-offs will be emphasized. Case studies will primarily be devoted to NAS for image classification. Finally, the open challenges of this popular research area will be discussed.
deep neural network, evolutionary algorithm, hardware accelerator, inference, image classification
@inbook{BUT185298,
author="Lukáš {Sekanina} and Vojtěch {Mrázek} and Michal {Piňos}",
title="Hardware-Aware Evolutionary Approaches to Deep Neural Networks",
booktitle="Handbook of Evolutionary Machine Learning",
year="2023",
publisher="Springer Nature Singapore",
address="Singapore",
series="Genetic and Evolutionary Computation",
pages="367--396",
doi="10.1007/978-981-99-3814-8\{_}12",
isbn="978-981-9938-13-1",
url="https://link.springer.com/chapter/10.1007/978-981-99-3814-8_12"
}