In this paper, we evaluate the reliability of the You Only Look Once (YOLO) object detection framework. We have exposed to controlled neutron beams GPUs designed with three different architectures (Kepler, Maxwell, and Pascal) running Darknet, a Convolutional Neural Network for automotive applications, detecting objects in both Caltech and Visual Object Classes data sets. By analyzing the neural network corrupted output, we can distinguish between tolerable errors and critical errors, i.e., errors that could impact on real-time system execution.Additionally, we propose an Algorithm-Based Fault-Tolerance (ABFT) strategy to apply to the matrix multiplication kernels of neural networks able to detect and correct 50% to 60% of radiation-induced corruptions. We experimentally validate our hardening solution and compare its efficiency and efficacy with the available ECC.
Evaluation and Mitigation of Soft-Errors in Neural Network-Based Object Detection in Three GPU Architectures / Fernandes dos Santos, Fernando; Draghetti, Lucas; Weigel, Lucas; Carro, Luigi; Navaux, Philippe; Rech, Paolo. - (2017), pp. 169-176. (Intervento presentato al convegno 2017 47th Annual IEEE/IFIP International Conference on Dependable Systems and Networks Workshops (DSN-W) tenutosi a Denver, Colorado nel 26th–29th June 2017) [10.1109/dsn-w.2017.47].
Evaluation and Mitigation of Soft-Errors in Neural Network-Based Object Detection in Three GPU Architectures
Rech, PaoloUltimo
2017-01-01
Abstract
In this paper, we evaluate the reliability of the You Only Look Once (YOLO) object detection framework. We have exposed to controlled neutron beams GPUs designed with three different architectures (Kepler, Maxwell, and Pascal) running Darknet, a Convolutional Neural Network for automotive applications, detecting objects in both Caltech and Visual Object Classes data sets. By analyzing the neural network corrupted output, we can distinguish between tolerable errors and critical errors, i.e., errors that could impact on real-time system execution.Additionally, we propose an Algorithm-Based Fault-Tolerance (ABFT) strategy to apply to the matrix multiplication kernels of neural networks able to detect and correct 50% to 60% of radiation-induced corruptions. We experimentally validate our hardening solution and compare its efficiency and efficacy with the available ECC.File | Dimensione | Formato | |
---|---|---|---|
Evaluation_and_Mitigation_of_Soft-Errors_in_Neural_Network-Based_Object_Detection_in_Three_GPU_Architectures.pdf
Solo gestori archivio
Tipologia:
Versione editoriale (Publisher’s layout)
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
584.41 kB
Formato
Adobe PDF
|
584.41 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione