☆ 4.5 Article

Designing Efficient DNNs via Hardware-Aware Neural Architecture Search and Beyond

IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS (2022)

Related references

Note: Only part of the references are listed.

Article Computer Science, Artificial Intelligence

Bringing AI to edge: From deep learning's perspective

Di Liu et al.

Summary: The intersection of edge computing and artificial intelligence is leading to the development of edge intelligence systems but faces challenges such as computational gaps. To bridge this gap, new techniques like lightweight deep learning models and neural architecture search have been proposed. A holistic review of these deep learning techniques is important for understanding the state-of-the-art in edge intelligence implementation.

NEUROCOMPUTING (2022)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

HSCoNAS: Hardware-Software Co-Design of Efficient DNNs via Neural Architecture Search

Xiangzhong Luo et al.

Summary: This paper introduces a novel hardware-aware neural architecture search framework, HSCoNAS, designed to automate the creation of deep neural networks with high accuracy and low latency on target hardware. The framework includes hardware performance modeling, dynamic channel scaling, and progressive space shrinking techniques to facilitate fine-grained and efficient explorations. Extensive experiments conducted on ImageNet demonstrate the superiority of HSCoNAS over recent state-of-the-art approaches on diverse target hardware such as GPU, CPU, and edge devices.

PROCEEDINGS OF THE 2021 DESIGN, AUTOMATION & TEST IN EUROPE CONFERENCE & EXHIBITION (DATE 2021) (2021)

Add to Collection

Article Computer Science, Hardware & Architecture