Enhancing Fault Resilience of QNNs by Selective Neuron Splitting (2306.09973v1)
Abstract: The superior performance of Deep Neural Networks (DNNs) has led to their application in various aspects of human life. Safety-critical applications are no exception and impose rigorous reliability requirements on DNNs. Quantized Neural Networks (QNNs) have emerged to tackle the complexity of DNN accelerators, however, they are more prone to reliability issues. In this paper, a recent analytical resilience assessment method is adapted for QNNs to identify critical neurons based on a Neuron Vulnerability Factor (NVF). Thereafter, a novel method for splitting the critical neurons is proposed that enables the design of a Lightweight Correction Unit (LCU) in the accelerator without redesigning its computational part. The method is validated by experiments on different QNNs and datasets. The results demonstrate that the proposed method for correcting the faults has a twice smaller overhead than a selective Triple Modular Redundancy (TMR) while achieving a similar level of fault resiliency.
- D. Silver et al., “Mastering the game of go without human knowledge,” nature, vol. 550, no. 7676, pp. 354–359, 2017.
- S. Mozaffari et al., “Deep learning-based vehicle behavior prediction for autonomous driving applications: A review,” IEEE T-ITS, 2020.
- Y. Ibrahim et al., “Soft errors in dnn accelerators: A comprehensive review,” Microelectronics Reliability, vol. 115, p. 113969, 2020.
- M. Shafique et al., “Robust machine learning systems: Challenges, current trends, perspectives, and the road ahead,” IEEE Design & Test, vol. 37, no. 2, pp. 30–57, 2020.
- U. Zahid et al., “Fat: Training neural networks for reliable inference under hardware faults,” in 2020 IEEE International Test Conference (ITC). IEEE, 2020, pp. 1–10.
- N. Khoshavi et al., “Fiji-fin: A fault injection framework on quantized neural network inference accelerator,” in 2020 19th IEEE International Conference on Machine Learning and Applications (ICMLA). IEEE, 2020, pp. 1139–1144.
- S. Mittal, “A survey on modeling and improving reliability of dnn algorithms and accelerators,” Journal of Systems Architecture, vol. 104, p. 101689, 2020.
- A. Mahmoud et al., “Hardnn: Feature map vulnerability evaluation in cnns,” arXiv preprint arXiv:2002.09786, 2020.
- C. Schorn and other, “Accurate neuron resilience prediction for a flexible reliability management in neural network accelerators,” in 2018 DATE. IEEE, 2018, pp. 979–984.
- C. Schorn et al., “An efficient bit-flip resilience optimization method for deep neural networks,” in 2019 DATE. IEEE, 2019, pp. 1507–1512.
- A. Ruospo and E. Sanchez, “On the reliability assessment of artificial neural networks running on ai-oriented mpsocs,” Applied Sciences, vol. 11, no. 14, p. 6455, 2021.
- M. Abdullah Hanif and M. Shafique, “Salvagednn: salvaging deep neural network accelerators with permanent faults through saliency-driven fault-aware mapping,” Philosophical Transactions of the Royal Society A, vol. 378, no. 2164, p. 20190164, 2020.
- M. H. Ahmadilivani et al., “Deepvigor: Vulnerability value ranges and factors for dnns reliability assessment,” in 28th IEEE European Test Symposium. In press, 2023.
- E. Ozen and A. Orailoglu, “Just say zero: Containing critical bit-error propagation in deep neural networks with anomalous feature suppression,” in 39th ICCAD, 2020, pp. 1–9.
- R. David et al., “Tensorflow lite micro: Embedded machine learning for tinyml systems,” Machine Learning and Systems, vol. 3, pp. 800–811, 2021.
- R. Leveugle et al., “Statistical fault injection: Quantified error and confidence,” in 2009 DATE. IEEE, 2009, pp. 502–506.
- Mohammad Hasan Ahmadilivani (9 papers)
- Mahdi Taheri (17 papers)
- Jaan Raik (26 papers)
- Masoud Daneshtalab (24 papers)
- Maksim Jenihhin (31 papers)