In recent years, Deep Neural Networks have been increasingly adopted by a wide range of applications characterized by high-reliability requirements, such as aerospace and automotive. In this paper, we propose an FPGA-based platform for emulating faults in the architecture of DNNs. The approach exploits the reconfigurability of FPGAs to mimic faults affecting the hardware implementing DNNs. The platform allows the emulation of various kinds of fault models enabling the possibility to adapt to different types, devices, and architectures. In this work, a fault injection campaign has been performed on a convolutional layer of AlexNet, demonstrating the feasibility of the platform. Furthermore, the errors induced in the layer are analyzed with respect to the impact on the whole network inference classification.

An Emulation Platform for Evaluating the Reliability of Deep Neural Networks / DE SIO, Corrado; Azimi, Sarah; Sterpone, Luca. - ELETTRONICO. - (2020), pp. 1-4. (Intervento presentato al convegno 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems (DFT 2020) tenutosi a ita nel 19-21 Oct. 2020) [10.1109/DFT50435.2020.9250872].

An Emulation Platform for Evaluating the Reliability of Deep Neural Networks

Corrado De Sio;Sarah Azimi;Luca Sterpone
2020

Abstract

In recent years, Deep Neural Networks have been increasingly adopted by a wide range of applications characterized by high-reliability requirements, such as aerospace and automotive. In this paper, we propose an FPGA-based platform for emulating faults in the architecture of DNNs. The approach exploits the reconfigurability of FPGAs to mimic faults affecting the hardware implementing DNNs. The platform allows the emulation of various kinds of fault models enabling the possibility to adapt to different types, devices, and architectures. In this work, a fault injection campaign has been performed on a convolutional layer of AlexNet, demonstrating the feasibility of the platform. Furthermore, the errors induced in the layer are analyzed with respect to the impact on the whole network inference classification.
2020
978-1-7281-9457-8
File in questo prodotto:
File Dimensione Formato  
DFT_CameaReady_FinalVersion.pdf

accesso aperto

Descrizione: Camera Ready-DFT 2020
Tipologia: 2. Post-print / Author's Accepted Manuscript
Licenza: PUBBLICO - Tutti i diritti riservati
Dimensione 537.54 kB
Formato Adobe PDF
537.54 kB Adobe PDF Visualizza/Apri
09250872.pdf

non disponibili

Tipologia: 2a Post-print versione editoriale / Version of Record
Licenza: Non Pubblico - Accesso privato/ristretto
Dimensione 707.47 kB
Formato Adobe PDF
707.47 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2844426