ABSTRACT
Deep Neural Networks (DNNs) have gained growing attention in many domain-specific supervised learning applications. However, the current DNNs still face two challenges. One is the difficulty of obtaining well-labeled training data for supervised learning and the other is concerned with the efficiency of training due to the lack of precise characteristics of the objects in the training process. We propose a framework of formal specification-based data generation for the training and testing of DNNs. The framework is characterized by using formal specifications to define the important and distinct features of the objects to be identified. The features are expected to serve as the foundation for generating training and testing data for DNNs. In this paper, we discuss all the activities involved in the framework and the detailed approach to writing the formal specifications. We also conduct a case study on traffic sign recognition to validate the framework.
- Jean-Raymond Abrial, Michael Butler, Stefan Hallerstede, Thai Son Hoang, Farhad Mehta, and Laurent Voisin. 2010. Rodin: an open toolset for modelling and reasoning in Event-B. International journal on software tools for technology transfer 12, 6 (2010), 447–466.Google ScholarCross Ref
- Gérard Berry. 2007. SCADE: Synchronous design and validation of embedded control software. In Next Generation Design and Verification Methodologies for Distributed Embedded Control Systems. Springer, 19–33.Google Scholar
- Sumathi Gokulanathan, Alexander Feldsher, Adi Malca, Clark Barrett, and Guy Katz. 2020. Simplifying neural networks using formal verification. In NASA Formal Methods Symposium. Springer, 85–93.Google ScholarDigital Library
- Gaetan JDR Hains, Arvid Jakobsson, and Youry Khmelevsky. 2019. Formal methods and software engineering for DL. Security, safety and productivity for DL systems development. arXiv preprint arXiv:1901.11334 (2019).Google Scholar
- Julia Hirschberg and Christopher D Manning. 2015. Advances in natural language processing. Science 349, 6245 (2015), 261–266.Google Scholar
- Sebastian Houben, Johannes Stallkamp, Jan Salmen, Marc Schlipsing, and Christian Igel. 2013. Detection of traffic signs in real-world images: The German Traffic Sign Detection Benchmark. In The 2013 international joint conference on neural networks (IJCNN). Ieee, 1–8.Google Scholar
- Licheng Jiao and Jin Zhao. 2019. A survey on the new generation of deep learning in image processing. IEEE Access 7 (2019), 172231–172263.Google ScholarCross Ref
- Cliff B Jones. 1990. Systematic software development using VDM. Prentice Hall International Series in Computer Science (1990).Google ScholarDigital Library
- Axel van Lamsweerde. 2000. Formal specification: a roadmap. In Proceedings of the Conference on the Future of Software Engineering. 147–159.Google ScholarDigital Library
- Ming Liang and Xiaolin Hu. 2015. Recurrent convolutional neural network for object recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition. 3367–3375.Google ScholarCross Ref
- JM Lillo-Castellano, I Mora-Jiménez, Carlos Figuera-Pozuelo, and José Luis Rojo-Álvarez. 2015. Traffic sign segmentation and classification using statistical learning methods. Neurocomputing 153 (2015), 286–299.Google ScholarCross Ref
- Shaoying Liu, A Jefferson Offutt, Chris Ho-Stuart, Yong Sun, and Mitsuru Ohba. 1998. SOFL: A formal engineering methodology for industrial applications. IEEE Transactions on Software Engineering 24, 1 (1998), 24–45.Google ScholarDigital Library
- Andreas Møgelmose. 2015. Visual analysis in traffic & re-identification. (2015).Google Scholar
- Shehla Riaz, Hamra Afzaal, Muhammad Imran, Nazir Ahmad Zafar, and Mehmet Sabih Aksoy. 2015. Formalizing mobile ad hoc and sensor networks Using VDM-SL. Procedia Computer Science 63 (2015), 148–153.Google ScholarCross Ref
- Sebastian Ruder. 2016. An overview of gradient descent optimization algorithms. arXiv preprint arXiv:1609.04747 (2016).Google Scholar
- Stuart Russell, Daniel Dewey, and Max Tegmark. 2015. Research priorities for robust and beneficial artificial intelligence. Ai Magazine 36, 4 (2015), 105–114.Google ScholarDigital Library
- Sanjit A Seshia, Ankush Desai, Tommaso Dreossi, Daniel J Fremont, Shromona Ghosh, Edward Kim, Sumukh Shivakumar, Marcell Vazquez-Chanlatte, and Xiangyu Yue. 2018. Formal specification for deep neural networks. In International Symposium on Automated Technology for Verification and Analysis. Springer, 20–34.Google ScholarCross Ref
- Sanjit A Seshia, Dorsa Sadigh, and S Shankar Sastry. 2022. Toward verified artificial intelligence. Commun. ACM 65, 7 (2022), 46–55.Google ScholarDigital Library
- Amanpreet Singh, Narina Thakur, and Aakanksha Sharma. 2016. A review of supervised machine learning algorithms. In 2016 3rd International Conference on Computing for Sustainable Global Development (INDIACom). Ieee, 1310–1315.Google Scholar
- J Michael Spivey. 1988. Understanding Z: a specification language and its formal semantics. Vol. 3. Cambridge University Press.Google Scholar
- Johannes Stallkamp, Marc Schlipsing, Jan Salmen, and Christian Igel. 2012. Man vs. computer: Benchmarking machine learning algorithms for traffic sign recognition. Neural networks 32 (2012), 323–332.Google Scholar
- Nikko Strom. 2015. Scalable distributed DNN training using commodity GPU cloud computing. In Sixteenth annual conference of the international speech communication association.Google Scholar
- Muhammad Usama, Junaid Qadir, Aunn Raza, Hunain Arif, Kok-Lim Alvin Yau, Yehia Elkhatib, Amir Hussain, and Ala Al-Fuqaha. 2019. Unsupervised machine learning for networking: Techniques, applications and research challenges. IEEE access 7 (2019), 65579–65615.Google Scholar
- Jim Woodcock, Peter Gorm Larsen, Juan Bicarregui, and John Fitzgerald. 2009. Formal methods: Practice and experience. ACM computing surveys (CSUR) 41, 4 (2009), 1–36.Google Scholar
- Xiaoyong Yuan, Pan He, Qile Zhu, and Xiaolin Li. 2019. Adversarial examples: Attacks and defenses for deep learning. IEEE transactions on neural networks and learning systems 30, 9 (2019), 2805–2824.Google Scholar
- Fatin Zaklouta and Bogdan Stanciulescu. 2012. Real-time traffic-sign recognition using tree classifiers. IEEE Transactions on Intelligent Transportation Systems 13, 4 (2012), 1507–1514.Google ScholarDigital Library
- Xiaojin Jerry Zhu. 2005. Semi-supervised learning literature survey. (2005).Google Scholar
- Yingying Zhu, Chengquan Zhang, Duoyou Zhou, Xinggang Wang, Xiang Bai, and Wenyu Liu. 2016. Traffic sign detection and recognition using fully convolutional network guided proposals. Neurocomputing 214 (2016), 758–766.Google ScholarDigital Library
Index Terms
- A Framework of Formal Specification-Based Data Generation for Deep Neural Networks
Recommendations
A Formal Framework for ASTRAL Intralevel Proof Obligations
ASTRAL is a formal specification language for real-time systems. It is intended to support formal software development, and therefore has been formally defined. This paper focuses on how to formally prove the mathematical correctness of ASTRAL ...
Design and Implementation of a Tool for Specifying Specification in SOFL
Revised Selected Papers of the Second International Workshop on Structured Object-Oriented Formal Language and Method - Volume 7787Structure Object-oriented Formal Language SOFL is not just a formal language for writing formal specification. It is also an approach and a methodology. SOFL provides a three-step approach for modelling a software system using formal specification. ...
A formal requirements engineering method for specification, synthesis, and verification
SEE '97: Proceedings of the 8th International Conference on Software Engineering Environments (SEE '97)This paper presents a formal requirements engineering method capturing specification, synthesis, and verification. Being multi-paradigm, our approach integrates individual established formal methods: temporal logics are used to express abstract ...
Comments