Evaluation of Histogram of Oriented Gradients Soft Errors Criticality for Automotive Applications
Open accessThe DOI matches its Crossref record; 47 of the 48 references it cites were found in a registry or an index.
Abstract
Pedestrian detection reliability is a key problem for autonomous or aided driving, and methods that use Histogram of Oriented Gradients (HOG) are very popular. Embedded Graphics Processing Units (GPUs) are exploited to run HOG in a very efficient manner. Unfortunately, GPUs architecture has been shown to be particularly vulnerable to radiation-induced failures. This article presents an experimental evaluation and analytical study of HOG reliability. We aim at quantifying and qualifying the radiation-induced errors on pedestrian detection applications executed in embedded GPUs. We analyze experimental results obtained executing HOG on embedded GPUs from two different vendors, exposed for about 100 hours to a controlled neutron beam at Los Alamos National Laboratory. We consider the number and position of detected objects as well as precision and recall to discriminate critical erroneous computations. The reported analysis shows that, while being intrinsically resilient (65% to 85% of output errors only slightly impact detection), HOG experienced some particularly critical errors that could result in undetected pedestrians or unnecessary vehicle stops. Additionally, we perform a fault-injection campaign to identify HOG critical procedures. We observe that Resize and Normalize are the most sensitive and critical phases, as about 20% of injections generate an output error that significantly impacts HOG detection. With our insights, we are able to find those limited portions of HOG that, if hardened, are more likely to increase reliability without introducing unnecessary overhead.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“radiation test offers the most adherent solution to cause soft errors, however, apart from the highly expensive equipment and the complex setup, there is little to no controllability on the injection process, and the observability is only on the primary outputs. As a result, it is suitable for final validation and blackbox campaigns, not for in-depth analysis and investigations [9], [36].”
From Fast and Accurate Error Simulation for CNNs Against Soft Errors (2023) · cites this paper as [36]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Histograms of Oriented Gradients for Human Detection
Dalal, N., & Triggs, B. (2005) · 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) · doi:10.1109/cvpr.2005.177 · Cited by 33,584 · cited 7×
“The method is composed of five GPU steps: Gradient computation, Orientation binning, Descriptor Blocks and Block normalization, and Classifier [Dalal and Triggs 2005].”
BackgroundMethodInfluentialAn introduction to ROC analysis
Fawcett, T. (2006) · Pattern Recognition Letters · doi:10.1016/j.patrec.2005.10.010 · Cited by 21,879 · cited 4×
“5 still holds, as the T J tradeoff discussion presented in Fawcett [2006] is valid independently from the source of detection imprecision (intrinsic algorithm detection imprecision or radiation-induced corruption).”
BackgroundMethodInfluentialAdvances in kernel methods: support vector learning
Schölkopf, B., Burges, C. J. C., & Smola, A. J. (1999) · doi:10.5555/299094 · Cited by 5,789 · cited 2×
“The soft linear SVM classifier is trained with SVMLight, which solves several problems on SVM vectors classifiers [Joachims 1999].”
MethodPedestrian Detection: An Evaluation of the State of the Art
Dollar, P., Wojek, C., Schiele, B., & Perona, P. (2012) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2011.155 · Cited by 3,190 · cited 1×
“…choice was made for two main reasons: (1) HOG combined with SVM is one of the most popular and effective detection systems [Dalal and Triggs 2005; Dollar et al. 2012] and (2) the availability of HOG together with SVM on a publicly accessible library eases the reproduction of our results [Bradski…”
ResultRadiation-induced soft errors in advanced semiconductor technologies
Baumann, R. C. (2005) · IEEE Transactions on Device and Materials Reliability · doi:10.1109/tdmr.2005.853449 · Cited by 1,549 · cited 5×
“The larger the cross section, the higher the use of sensitive resources [Baumann 2005; Mukherjee et al. 2003].”
BackgroundInfluentialFast Human Detection Using a Cascade of Histograms of Oriented Gradients
Zhu, Q., Yeh, M.-C., Cheng, K.-T., & Avidan, S. (2006) · 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2 (CVPR'06) · doi:10.1109/cvpr.2006.119 · Cited by 1,399 · cited 2×
“We specifically consider the Histograms of Oriented Gradients (HOG) algorithm [Dalal and Triggs 2005], which is the core of several pedestrian or object detection systems [Sivaraman and Trivedi 2013; Zhu et al. 2006; Kang and Lim 2014; Li and Guo 2013].”
BackgroundMethodSivaraman, S., & Trivedi, M. M. (2013) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2013.2266661 · Cited by 908 · cited 2×
“We specifically consider the Histograms of Oriented Gradients (HOG) algorithm [Dalal and Triggs 2005], which is the core of several pedestrian or object detection systems [Sivaraman and Trivedi 2013; Zhu et al. 2006; Kang and Lim 2014; Li and Guo 2013].”
BackgroundMethodComputer Architecture, Fifth Edition: A Quantitative Approach
Hennessy, J. L., & Patterson, D. A. (2011) · Morgan Kaufmann Publishers Inc. eBooks · Cited by 880
Mukherjee, S. S., Weaver, C. T., Emer, J., Reinhardt, S. K., & Austin, T. M. (2003) · Cited by 840
Recognizing actions using depth motion maps-based histograms of oriented gradients
Yang, X., Zhang, C., & Tian, Y. (2012) · Proceedings of the 20th ACM international conference on Multimedia · doi:10.1145/2393347.2396382 · Cited by 564 · cited 1×
“HOG is one of the most common features for pattern or object detection [Yang et al. 2012; Lim et al. 2013; Ren and Ramanan 2013], particularly in automotive Dalal and Triggs [2005]. applications [Sivaraman and Trivedi 2013], and its application on both human [Zhu et al. 2006; Kang and Lim 2014] and…”
Background
Research works printed but not linked: 10
- G. Bradski. 2000. The open source computer vision (OpenCV) library.
- Dr. Dobb's J. Softw. Tools 1, 1 (2000), 1.
- 10.5555/647883.738252 · doi:10.5555/647883.738252
- Google. 2016a. Google Self-Driving Car Project Montly Report. Retrieved June 1, 2016, from https://www. google.com/selfdrivingcar/reports/.
- S. Hari, R. Venkatagiri, S. Adve, and H. Naeimi. 2014. TGangES: Gang error simulation for hardware resiliency evaluation. In Proceeding of the 41st Annual International Symposium on Computer Archite- cuture (ISCA'14).
- 10.5555/1899254.1899261 · doi:10.5555/1899254.1899261
- Subhasish Mitra. 2010. System-Level Single-Event Effects. IEEE Nuclear and Space Radiation Effects Conference, NSREC 2012 Short Course. (2010).
- 10.5555/832299.836499 · doi:10.5555/832299.836499
- Nvidia. 2014. Tegra K1 techinical reference manual. DP-06905-001-v03p (2014).
- Nvidia. 2016. Tegra X1 techinical reference manual. TRM DP 07225 001-v1.1p (2016).
Cited by
Understanding error propagation in deep learning neural network (DNN) accelerators and applications
Li, G., Hari, S. K. S., Sullivan, M., Tsai, T., Pattabiraman, K., Emer, J., & Keckler, S. W. (2017) · Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis · doi:10.1145/3126908.3126964 · Cited by 492
“[21] evaluated the resilience of histogram of oriented gradient applications for self-driving cars, but they did not consider DNNs.”
BackgroundKernel and layer vulnerability factor to evaluate object detection reliability in GPUs
Fernandes dos Santos, F., Carro, L., & Rech, P. (2019) · IET Computers & Digital Techniques · doi:10.1049/iet-cdt.2018.5026 · Cited by 27
“For an in-depth discussion and analysis of the observed results, please refer to [27].”
BackgroundSelf-citationDetecting Errors in Convolutional Neural Networks Using Inter Frame Spatio-Temporal Correlation
Draghetti, L. K., dos Santos, F. F., Carro, L., & Rech, P. (2019) · IEEE International On-Line Testing Symposium (IOLTS) · doi:10.1109/iolts.2019.8854431 · Cited by 24
“However, while autonomous cars are expected to reduce of two-three orders of magnitude the number of accidents [1], [2], current self-driving systems are not yet ready to be employed in large-scale.”
BackgroundSelf-citationTensorFI+: A Scalable Fault Injection Framework for Modern Deep Learning Neural Networks
Laskar, S., Rahman, M. H., & Li, G. (2022) · IEEE International Symposium on Software Reliability Engineering Workshops (ISSREW) · doi:10.1109/issrew55968.2022.00074 · Cited by 15
“With the ever-increasing demands of DNNs, parallel processing units such as Graphics Processing Units (GPUs) and Tensor Processing Units (TPUs) have been developed to accelerate the DNN inference [3], [4].”
MethodInvestigating the impact of transient hardware faults on deep learning neural network inference
Rahman, M. H., Laskar, S., & Li, G. (2024) · Software Testing Verification and Reliability · doi:10.1002/stvr.1873 · Cited by 12
Metadata
- Authors
- Fernandes, F., Weigel, L., Jung, C., Navaux, P., Carro, L., & Rech, P.
- Affiliations
- Federal University of Rio Grande do Sul, Agronomia, Porto Alegre - RS
- Universidade Federal do Rio Grande do Sul (BR)
- Venue
- ACM Transactions on Architecture and Code Optimization
- Published
- 2016-11-15
- Volume
- 13
- Issue
- 4
- Pages
- 1-25
- ISSN
- 1544-3566 (ISSN-L)
- 1544-3973
- OpenAlex
- W2555834960
- Cited by
- 14
- References
- 46
- Citations by year
- 2024: 1 · 2023: 1 · 2022: 4 · 2021: 1 · 2020: 1 · 2019: 2 · 2018: 1 · 2017: 3
- FWCI
- 0.92
- Type
- journal article
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- ACM Transactions on Architecture and Code Optimization (publishedVersion)
- Institutional Research Information System (Università degli Studi di Trento) (submittedVersion)
- Funders
- Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
- Conselho Nacional de Desenvolvimento Científico e Tecnológico
- Topics
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Radiation Effects in Electronics
- Physical Sciences › Computer Science › Computer Networks and Communications › Distributed systems and fault tolerance
- Physical Sciences › Computer Science › Hardware and Architecture › Real-Time Systems Scheduling
- Keywords
- histogram of oriented gradients, soft errors, radiation-induced faults, embedded GPU, pedestrian detection, autonomous driving, neutron irradiation, fault injection, criticality analysis, hardware reliability, fault tolerance
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.