{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,27]],"date-time":"2026-05-27T22:17:00Z","timestamp":1779920220236,"version":"3.53.1"},"reference-count":81,"publisher":"IEEE","license":[{"start":{"date-parts":[[2021,10,1]],"date-time":"2021-10-01T00:00:00Z","timestamp":1633046400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2021,10,1]],"date-time":"2021-10-01T00:00:00Z","timestamp":1633046400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2021,10]]},"DOI":"10.1109\/issre52982.2021.00025","type":"proceedings-article","created":{"date-parts":[[2022,2,11]],"date-time":"2022-02-11T20:01:25Z","timestamp":1644609685000},"page":"127-138","source":"Crossref","is-referenced-by-count":34,"title":["Optimizing Selective Protection for CNN Resilience"],"prefix":"10.1109","author":[{"given":"Abdulrahman","family":"Mahmoud","sequence":"first","affiliation":[{"name":"Harvard University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Siva Kumar","family":"Sastry Hari","sequence":"additional","affiliation":[{"name":"NVIDIA Corporation"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Christopher W.","family":"Fletcher","sequence":"additional","affiliation":[{"name":"University of Illinois at Urbana-Champaign"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sarita V.","family":"Adve","sequence":"additional","affiliation":[{"name":"University of Illinois at Urbana-Champaign"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Charbel","family":"Sakr","sequence":"additional","affiliation":[{"name":"NVIDIA Corporation"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Naresh","family":"Shanbhag","sequence":"additional","affiliation":[{"name":"University of Illinois at Urbana-Champaign"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Pavlo","family":"Molchanov","sequence":"additional","affiliation":[{"name":"NVIDIA Corporation"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael B.","family":"Sullivan","sequence":"additional","affiliation":[{"name":"NVIDIA Corporation"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Timothy","family":"Tsai","sequence":"additional","affiliation":[{"name":"NVIDIA Corporation"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Stephen W.","family":"Keckler","sequence":"additional","affiliation":[{"name":"NVIDIA Corporation"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref73","article-title":"Going deeper with convolutions","volume":"abs 1409 4842","author":"szegedy","year":"2014","journal-title":"CoRR"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.2200\/S01004ED1V01Y202004CAC050"},{"key":"ref71","author":"systems","year":"2012","journal-title":"Sample size calculator Website"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA45697.2020.00086"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2017.2742698"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.1109\/MICRO.2016.7783745"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2020.2975764"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1109\/QRS.2015.13"},{"key":"ref75","article-title":"Failure analysis and quantification for contemporary and future supercomputers","volume":"abs 1911 2118","author":"tan","year":"2019","journal-title":"ArXiv"},{"key":"ref38","article-title":"Dynamic deep neural networks: Optimizing accuracy-efficiency trade-offs by selective execution","volume":"abs 1701 299","author":"liu","year":"2017","journal-title":"CoRR"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1109\/ICCAD45719.2019.8942127"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.1109\/TDSC.2006.40"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1145\/3126908.3126964"},{"key":"ref32","first-page":"554","article-title":"Lessons learned from memory errors observed over the lifetime of cielo","author":"levy","year":"0","journal-title":"SC14 International Conference for High Performance Computing Networking Storage and Analysis SC"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/DATE.2009.5090716"},{"key":"ref30","doi-asserted-by":"crossref","first-page":"227","DOI":"10.1145\/2854038.2854059","article-title":"IPAS: Intelligent Protection Against Silent Output Corruption in Scientific Applications","author":"ignacio laguna","year":"2016","journal-title":"IEEE\/ACM International Symposium on Code Generation and Optimization (CGO) ACM"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/TNS.2018.2884460"},{"key":"ref36","article-title":"Enabling sparse winograd convolution by native pruning","volume":"abs 1702 8597","author":"li","year":"2017","journal-title":"ArXiv"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1145\/1346281.1346315"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/DSN.2018.00016"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-015-0816-y"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1109\/DSN.2008.4630072"},{"key":"ref61","article-title":"Toyota unintended acceleration and the big bowl of &#x2018;spaghetti&#x2019; code","year":"2013","journal-title":"Safety Research & Strategies Inc"},{"key":"ref63","article-title":"Analytical guarantees on numerical precision of deep neural networks","author":"sakr","year":"2017","journal-title":"ICML"},{"key":"ref28","article-title":"Exascale computing study: Technology challenges in achieving exascale systems","volume":"15","author":"kogge","year":"2008","journal-title":"Defense Advanced Research Projects Agency Information Processing Techniques Office (DARPA IPTO) Techinal Representative"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2018.8461702"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA52012.2021.00010"},{"key":"ref65","article-title":"Inverted residuals and linear bottlenecks: Mobile networks for classification, detection and segmentation","volume":"abs 1801 4381","author":"sandler","year":"2018","journal-title":"CoRR"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.23919\/DATE.2018.8342151"},{"key":"ref29","article-title":"One weird trick for parallelizing convolutional neural networks","volume":"abs 1404 5997","author":"krizhevsky","year":"2014","journal-title":"CoRR"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.23919\/DATE.2019.8714885"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1186\/s40537-019-0197-0"},{"key":"ref69","article-title":"Very deep convolutional networks for large-scale image recognition","volume":"abs 1409 1556","author":"simonyan","year":"2014","journal-title":"CoRR"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.32985\/ijeces.12.2.4"},{"key":"ref1","article-title":"Detecting out-of-distribution inputs in deep neural networks using an early-layer output","volume":"abs 1910 10307","author":"abdelzad","year":"2019","journal-title":"ArXiv"},{"key":"ref20","article-title":"Deep residual learning for image recognition","volume":"abs 1512 3385","author":"he","year":"2015","journal-title":"CoRR"},{"key":"ref22","first-page":"497","article-title":"Terminal brain damage: Exposing the graceless degradation in deep neural networks under hardware fault attacks","author":"hong","year":"2019","journal-title":"28th USENIX Security Symposium (USENIX Security 19)"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/MICRO50266.2020.00033"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2020.2968129"},{"key":"ref23","article-title":"Squeezenet: Alexnet-level accuracy with 50x fewer parameters and <1mb model size","volume":"abs 1602 7360","author":"iandola","year":"2016","journal-title":"CoRR"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00286"},{"key":"ref25","article-title":"Road vehicles - Functional safety","year":"2011","journal-title":"International Organization for Standardization"},{"key":"ref50","year":"2020","journal-title":"&#x201C;NVIDIA Data Center Deep Learning Product Performance"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1007\/s10836-020-05920-2"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1109\/ICIP.2018.8451268"},{"key":"ref58","article-title":"Pytorch classification models","year":"2019","journal-title":"PyTorch"},{"key":"ref57","first-page":"445","author":"ping","year":"2020","journal-title":"SERN Modeling and Analyzing the Soft Error Reliability of Convolutional Neural Networks"},{"key":"ref56","article-title":"The effectiveness of data augmentation in image classification using deep learning","volume":"abs 1712 4621","author":"perez","year":"2017","journal-title":"CoRR"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1109\/EDCC.2006.9"},{"key":"ref54","article-title":"Automatic differentiation in pytorch","author":"paszke","year":"2017","journal-title":"NIPS-W"},{"key":"ref53","first-page":"1","article-title":"Just say zero: Containing critical bit-error propagation in deep neural networks with anomalous feature suppression","year":"0","journal-title":"2020 IEEE\/ACM International Conference On Computer Aided Design (ICCAD)"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2020.3012209"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1145\/3295500.3356177"},{"key":"ref11","article-title":"Technical report: Nemo dnn quantization for deployment model","volume":"abs 2004 5930","author":"conti","year":"2020","journal-title":"ArXiv"},{"key":"ref40","article-title":"Shufflenet V2: practical guidelines for efficient CNN architecture design","volume":"abs 1807 11164","author":"ma","year":"2018","journal-title":"CoRR"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/TPDS.2018.2864184"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1145\/1736020.1736063"},{"key":"ref14","article-title":"A survey of quantization methods for efficient neural network inference","volume":"abs 2103 13630","author":"gholami","year":"2021","journal-title":"ArXiv"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/ISQED51717.2021.9424287"},{"key":"ref16","article-title":"In-place zero-space memory protection for cnn","volume":"abs 1910 14479","author":"guan","year":"2019","journal-title":"ArXiv"},{"key":"ref81","author":"zhang","year":"2020","journal-title":"Towards energy-efficient and reliable deep learning inference"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1145\/2150976.2150990"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/TDSC.2021.3063083"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/ISPASS.2017.7975296"},{"key":"ref80","article-title":"FSD Chip - Tesla WikiChip","year":"2019","journal-title":"Wikichip"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1145\/3005348"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2021.3076716"},{"key":"ref6","article-title":"Toward Exascale Resilience: 2014 Update","author":"cappello","year":"2014","journal-title":"International Journal on Supercomputing Frontiers and Innovations"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/2807591.2807670"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/DSN-W.2018.00046"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1177\/1094342009347767"},{"key":"ref49","article-title":"Improving INT8 Accuracy Using Quantization Aware Training and the NVIDIA Transfer Learning Toolkit","year":"2020","journal-title":"Nvidia"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/DSN48987.2021.00018"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1016\/j.sysarc.2019.101689"},{"key":"ref45","author":"miller","year":"2010","journal-title":"Amazon Web Services"},{"key":"ref48","article-title":"NVIDIA Tesla V100 GPU Accelerator","year":"2018","journal-title":"Nvidia"},{"key":"ref47","article-title":"Pruning convolutional neural networks for resource efficient transfer learning","volume":"abs 1611 6440","author":"molchanov","year":"2016","journal-title":"CoRR"},{"key":"ref42","article-title":"Hardnn: Feature map vulnerability evaluation in cnns","volume":"abs 2002 9786","author":"mahmoud","year":"2020","journal-title":"ArXiv"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1109\/DSN-W50199.2020.00014"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1145\/3297858.3304050"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/SC.2018.00070"}],"event":{"name":"2021 IEEE 32nd International Symposium on Software Reliability Engineering (ISSRE)","location":"Wuhan, China","start":{"date-parts":[[2021,10,25]]},"end":{"date-parts":[[2021,10,28]]}},"container-title":["2021 IEEE 32nd International Symposium on Software Reliability Engineering (ISSRE)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/9700160\/9700163\/09700317.pdf?arnumber=9700317","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,6,13]],"date-time":"2022-06-13T17:07:19Z","timestamp":1655140039000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/9700317\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,10]]},"references-count":81,"URL":"https:\/\/doi.org\/10.1109\/issre52982.2021.00025","relation":{},"subject":[],"published":{"date-parts":[[2021,10]]}}}