A Low-cost Fault Corrector for Deep Neural Networks through Range Restriction
- 7 October 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Design & Test
- Vol. PP (21682356) , 1
Abstract
Deep neural networks (DNNs) have seen growing adoption in safety-critical domains. Unfortunately, they are also subject to unexpected failures due to hardware transient faults (soft errors). Traditional fault tolerance techniques require significant implementation efforts and/or incur major performance overheads. This work introduces Ranger, a low-cost fault corrector that can directly correct the faulty prediction output due to transient faults without re-computation. This research laid the foundations of improving the fault tolerance of DNN applications under hardware transient faults and it has influenced subsequent work in the area, both in academia and industry.Keywords
This publication has 9 references indexed in Scilit:
- Strategic Resilience Evaluation of Neural Networks Within Autonomous Vehicle Software2024
- Dr. DNA: Combating Silent Data Corruptions in Deep Learning using Distribution of Neuron Activations2024
- A Low-Cost Strategic Monitoring Approach for Scalable and Interpretable Error Detection in Deep Neural Networks2023
- Evaluation and Mitigation of Faults Affecting Swin Transformers2023
- Error Resilient Transformers: A Novel Soft Error Vulnerability Guided Approach to Error Checking and Suppression2023
- Strategies for Improving the Error Robustness of Convolutional Neural Networks2022
- Improving Fault Tolerance for Reliable DNN Using Boundary-Aware ActivationIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2021
- A Low-cost Fault Corrector for Deep Neural Networks through Range Restriction2021
- Efficient On-Line Error Detection and Mitigation for Deep Neural Network Accelerators2018