nach oben

Erschienen in:

2019 | OriginalPaper | Buchkapitel

Open Questions in Testing of Learned Computer Vision Functions for Automated Driving

verfasst von : Matthias Woehrle, Christoph Gladisch, Christian Heinzemann

Erschienen in: Computer Safety, Reliability, and Security

Verlag: Springer International Publishing

Einloggen

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config

KI-gestützte Suche

Aus

Abstract

Vision is an important sensing modality in automated driving. Deep learning-based approaches have gained popularity for different computer vision (CV) tasks such as semantic segmentation and object detection. However, the black-box nature of deep neural nets (DNN) is a challenge for practical software verification. With this paper, we want to initiate a discussion in the academic community about research questions w.r.t. software testing of DNNs for safety-critical CV tasks. To this end, we provide an overview of related work from various domains, including software testing, machine learning and computer vision and derive a set of open research questions to start discussion between the fields.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

über 102.000 Bücher
über 537 Zeitschriften

aus folgenden Fachgebieten:

Automobil + Motoren
Bauwesen + Immobilien
Business IT + Informatik
Elektrotechnik + Elektronik
Energie + Nachhaltigkeit
Finance + Banking
Management + Führung
Marketing + Vertrieb
Maschinenbau + Werkstoffe
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Springer Professional "Technik"

Online-Abonnement

Mit Springer Professional "Technik" erhalten Sie Zugriff auf:

über 67.000 Bücher
über 390 Zeitschriften

aus folgenden Fachgebieten:

Automobil + Motoren
Bauwesen + Immobilien
Business IT + Informatik
Elektrotechnik + Elektronik
Energie + Nachhaltigkeit
Maschinenbau + Werkstoffe

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

über 67.000 Bücher
über 340 Zeitschriften

aus folgenden Fachgebieten:

Bauwesen + Immobilien
Business IT + Informatik
Finance + Banking
Management + Führung
Marketing + Vertrieb
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Vorheriges Kapitel A Safety Standard Approach for Fully Autonomous Vehicles

Nächstes Kapitel Adaptive Deployment of Safety Monitors for Autonomous Systems

Many verification techniques are up to now only applied to image classification. While this simpler CV task is not relevant for our application, the corresponding methods are good starting points for further study.

We discuss first steps in this direction in the context of synthetic data in Sec. 2.3.

With the discriminative model we could only dismiss generated irrelevant test inputs, while with a generative model, we could directly generate relevant tests.

AIT Austrian Institute of Technology GmbH: CV-HAZOP VITRO. https://vitro-testing.com/cv-hazop/. Accessed 28 Mar 2019

Alhaija, H.A., Mustikovela, S.K., Mescheder, L.M., Geiger, A., Rother, C.: Augmented reality meets computer vision: efficient data generation for urban driving scenes. J. Comput. Vis. 126(9), 961–972 (2018)CrossRef

Barr, E.T., Harman, M., McMinn, P., Shahbaz, M., Yoo, S.: The oracle problem in software testing: a survey. IEEE Trans. Software Eng. 41(5), 507–525 (2015)CrossRef

Borg, M., et al.: Safely entering the deep: a review of verification and validation for machine learning and a challenge elicitation in the automotive industry. arXiv preprint arXiv:1812.05389 (2018)

Braiek, H.B., Khomh, F.: On testing machine learning programs. CoRR arXiv:abs/1812.02257 (2018)

Carlson, A., Skinner, K.A., Vasudevan, R., Johnson-Roberson, M.: Sensor transfer: learning optimal sensor effect image augmentation for sim-to-real domain adaptation. IEEE Robot. Autom. Lett. 4(3), 2431–2438 (2019)

Cheng, C.H., Huang, C.H., Brunner, T., Hashemi, V.: Towards safety verification of direct perception neural networks. arXiv preprint arXiv:1904.04706 (2019)

Cordts, M., et al.: The cityscapes dataset for semantic urban scene understanding. In: CVPR 2016 (2016)

Deng, J., Dong, W., Socher, R., Li, L., Li, K., Li, F.: Imagenet: a large-scale hierarchical image database. In: IEEE CVPR 2009, pp. 248–255 (2009)

10.

Franke, U.: 30 years fighting for robustness, June 2018. http://www.robustvision.net. Talk at Robust Vision Challenge (2018)

11.

Frénay, B., Kabán, A.: A comprehensive introduction to label noise. In: ESANN 2014 (2014)

12.

Frtunikj, J., Fuerst, S.: Engineering safe machine learning for automated driving systems. In: 27th Safety-Critical Systems Symposium (2019)

13.

Geiger, A., Lenz, P., Urtasun, R.: Are we ready for autonomous driving? The KITTI vision benchmark suite. In: IEEE CVPR 2012, pp. 3354–3361 (2012)

14.

Goodfellow, I.J., Shlens, J., Szegedy, C.: Explaining and harnessing adversarial examples. In: ICLR (2015)

15.

Group, W.A.I.Q.W.: Standard for automotive system image quality. In: IEEE P2020, IEEE (2019)

16.

Hendrycks, D., Dietterich, T.: Benchmarking neural network robustness to common corruptions and perturbations. arXiv preprint arXiv:1903.12261 (2019)

17.

Hutter, A.: Einsatz von Simulationsmodellen beim Test elektronischer Steuergeräte. In: Sax, E. (ed.) Automatisiertes Testen Eingebetteter Systeme in der Automobilindustrie. Hanser (2008)

18.

Jung, A.: Image augmentation for machine learning experiments (2019). https://github.com/aleju/imgaug. Accessed 9 Apr 2019

19.

Kar, A., et al.: Meta-Sim: Learning to Generate Synthetic Datasets. arXiv e-prints arXiv:1904.11621, April 2019

20.

Karpathy, A.: A recipe for training neural networks, April 2019. http://karpathy.github.io/2019/04/25/recipe/. blog Accessed 2019 Apr 30

21.

Koopman, P., Fratrik, F.: How many operational design domains, objects, and events? In: Workshop on AI Safety @ AAAI 2019 (2019)

22.

Koopman, P., Wagner, M.: Toward a framework for highly automated vehicle safety validation. Technical report, SAE Technical Paper (2018)

23.

Liu, C., Arnon, T., Lazarus, C., Barrett, C., Kochenderfer, M.J.: Algorithms for verifying deep neural networks. CoRR arXiv:abs/1903.06758 (2019)

24.

Mayer, N., et al.: What makes good synthetic training data for learning disparity and optical flow estimation? Int. J. Comput. Vis. 126(9), 942–960 (2018)CrossRef

25.

Meyer, B.: Soundness and completeness: with precision, April 2019. https://bertrandmeyer.com/2019/04/21/soundness-completeness-precision. blog Accessed 02 May 2019

26.

Pei, K., Cao, Y., Yang, J., Jana, S.: DeepXplore: automated whitebox testing of deep learning systems. In: Proceedings of SOSP 2017, pp. 1–18 (2017)

27.

Pezzementi, Z., et al.: Putting image manipulations in context: robustness testing for safe perception. In: 2018 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR), pp. 1–8 (2018)

28.

Poibrenski, A., Sprenger, J., Muller, C.: Toward a methodology for training with synthetic data on the example of pedestrian detection in a frame-by-frame semantic segmentation task. In: SEFAIAS@ICSE 2018, pp. 31–34 (2018)

29.

Qin, C., et al.: Verification of non-linear specifications for neural networks. CoRR arXiv:abs/1902.09592 (2019)

30.

Ré, C.: Software 2.0 and snorkel: beyond hand-labeled data. In: Proceedings of 24th ACM KDD 2018, 19–23 August 2018, p. 2876 (2018)

31.

SCSC: Data safety guidance. SCSC Version 3.1, The Safety-Critical Systems Club, York, Great Britain (2019)

32.

Sen, K., Marinov, D., Agha, G.: CUTE: a concolic unit testing engine for c. In: ACM SIGSOFT Software Engineering Notes, vol. 30, pp. 263–272. ACM (2005)

33.

Seshia, S.A., et al.: Formal specification for deep neural networks. In: Lahiri, S.K., Wang, C. (eds.) ATVA 2018. LNCS, vol. 11138, pp. 20–34. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-01090-4_2CrossRef

34.

Shetty, R., Schiele, B., Fritz, M.: Not using the car to see the sidewalk: quantifying and controlling the effects of context in classification and segmentation. CoRR arXiv:abs/1812.06707 (2018)

35.

Singh, G., Gehr, T., Püschel, M., Vechev, M.T.: An abstract domain for certifying neural networks. PACMPL 3(POPL), 41:1–41:30 (2019)

36.

Sun, Y., Wu, M., Ruan, W., Huang, X., Kwiatkowska, M., Kroening, D.: Concolic testing for deep neural networks. In: Proceedings of ASE 2018, pp. 109–119 (2018)

37.

Tian, Y., Pei, K., Jana, S., Ray, B.: DeepTest: automated testing of deep-neural-network-driven autonomous cars. In: arXiv:1708.08559 (2017)

38.

Wong, E., Schmidt, F.R., Metzen, J.H., Kolter, J.Z.: Scaling provable adversarial defenses. In: NeurIPS 2018, pp. 8410–8419 (2018)

39.

Xie, X., Ho, J.W.K., Murphy, C., Kaiser, G.E., Xu, B., Chen, T.Y.: Testing and validating machine learning classifiers by metamorphic testing. J. Syst. Softw. 84(4), 544–558 (2011)CrossRef

40.

Yuille, A.L., Liu, C.: Deep nets: What have they ever done for vision? arXiv preprint arXiv:1805.04025 (2018)

41.

Zendel, O., Honauer, K., Murschitz, M., Steininger, D., Domínguez, G.F.: WildDash - creating hazard-aware benchmarks. In: Ferrari, V., Hebert, M., Sminchisescu, C., Weiss, Y. (eds.) ECCV 2018. LNCS, vol. 11210, pp. 407–421. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-01231-1_25CrossRef

42.

Zendel, O., Murschitz, M., Humenberger, M., Herzner, W.: CV-HAZOP: introducing test data validation for computer vision. In: ICCV 2015, pp. 2066–2074 (2015)

43.

Zhang, M., Zhang, Y., Zhang, L., Liu, C., Khurshid, S.: DeepRoad: GAN-based metamorphic testing and input validation framework for autonomous driving systems. In: ASE 2018, pp. 132–142 (2018)

44.

Zhang, Y., Qiu, W., Chen, Q., Hu, X., Yuille, A.L.: UnrealStereo: controlling hazardous factors to analyze stereo vision. In: Proceedings of 3DV 2018, pp. 228–237 (2018)

Titel: Open Questions in Testing of Learned Computer Vision Functions for Automated Driving
verfasst von: Matthias Woehrle
Christoph Gladisch
Christian Heinzemann
Verlag: Springer International Publishing
Buch: Computer Safety, Reliability, and Security
Print ISBN: 978-3-030-26249-5

Electronic ISBN: 978-3-030-26250-1

Copyright-Jahr: 2019
DOI: https://doi.org/10.1007/978-3-030-26250-1_27

Springer Professional

Abstract

Bitte loggen Sie sich ein, um Zugang zu Ihrer Lizenz zu erhalten.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Springer Professional "Technik"

Springer Professional "Wirtschaft"

Premium Partner