Publication detail

RoHNAS: A Neural Architecture Search Framework with Conjoint Optimization for Adversarial Robustness and Hardware Efficiency of Convolutional and Capsule Networks

MARCHISIO, A. MRÁZEK, V. MASSA, A. BUSSOLINO, B. MARTINA, M. SHAFIQUE, M.

Original Title

RoHNAS: A Neural Architecture Search Framework with Conjoint Optimization for Adversarial Robustness and Hardware Efficiency of Convolutional and Capsule Networks

Type

journal article in Web of Science

Language

English

Original Abstract

Neural Architecture Search (NAS) algorithms aim at finding efficient Deep Neural Network (DNN) architectures for a given application under given system constraints. DNNs are computationally-complex as well as vulnerable to adversarial attacks. In order to address multiple design objectives, we propose RoHNAS, a novel NAS framework that jointly optimizes for adversarial-robustness and hardware-efficiency of DNNs executed on specialized hardware accelerators. Besides the traditional convolutional DNNs, RoHNAS additionally accounts for complex types of DNNs such as Capsule Networks. For reducing the exploration time, RoHNAS analyzes and selects appropriate values of adversarial perturbation for each dataset to employ in the NAS flow. Extensive evaluations on multi - Graphics Processing Unit (GPU) - High Performance Computing (HPC) nodes provide a set of Pareto-optimal solutions, leveraging the tradeoff between the above-discussed design objectives. For example, a Pareto-optimal DNN for the CIFAR-10 dataset exhibits 86.07 % accuracy, while having an energy of 38.63 mJ, a memory footprint of 11.85 MiB, and a latency of 4.47 ms.

Keywords

Adversarial Robustness, Energy Efficiency, Latency, Memory, Hardware-Aware Neural Architecture Search, Evolutionary Algorithm, Deep Neural Networks, Capsule Networks

Authors

MARCHISIO, A.; MRÁZEK, V.; MASSA, A.; BUSSOLINO, B.; MARTINA, M.; SHAFIQUE, M.

Released

1. 10. 2022

ISBN

2169-3536

Periodical

IEEE Access

Year of study

2022

Number

10

State

United States of America

Pages from

109043

Pages to

109055

Pages count

13

URL

https://ieeexplore.ieee.org/document/9917535

BibTex

@article{BUT179460,
  author="MARCHISIO, A. and MRÁZEK, V. and MASSA, A. and BUSSOLINO, B. and MARTINA, M. and SHAFIQUE, M.",
  title="RoHNAS: A Neural Architecture Search Framework with Conjoint Optimization for Adversarial Robustness and Hardware Efficiency of Convolutional and Capsule Networks",
  journal="IEEE Access",
  year="2022",
  volume="2022",
  number="10",
  pages="109043--109055",
  doi="10.1109/ACCESS.2022.3214312",
  issn="2169-3536",
  url="https://ieeexplore.ieee.org/document/9917535"
}