May 9, 2o23. 3PM (Rescheduled from November 2022)
Room: TG23 (Town Hall Building)
Speaker: Gunel Jahangirova
Title: Mutation Testing of Deep Learning Systems
Abstract: Deep Learning (DL) is increasingly adopted to solve complex tasks such as image recognition or autonomous driving. Companies are considering the inclusion of DL components in production systems, but one of their main concerns is how to assess the quality of such systems. Mutation testing is a technique to inject artificial faults into a system, under the assumption that the capability to expose (kill) such artificial faults translates into the capability to expose also real faults. In this talk, I will provide an overview of our work in adapting the idea behind mutation testing to DL systems. First, I will cover our investigation of what is a fault in a DL system. Then, I will introduce DeepCrime which is a mutation testing tool for DL systems based on the set of real faults that we have collected. Lastly, I will go through the implemented applications of mutation testing to the various software testing tasks such as test input generation, fault localisation, program repair, and test oracle generation in the domain of autonomous systems.
Bio: Gunel Jahangirova is a Lecturer at King’s College London (KCL). Before joining KCL, she worked as a PostDoctoral Researcher at the Software Institute of USI in Lugano, Switzerland. She had her PhD in a joint program between Fondazione Bruno Kessler in Trento, Italy and University College London, UK during which she was advised by Prof. Paolo Tonella, Dr. David Clark and Prof. Mark Harman. Her PhD work focused on the oracle problem in software testing, in particular, the assessment, improvement and placement of test oracles. She has master’s degree in “Management and Information Systems” from the University of Manchester and master’s and bachelor’s degree in “Informatics” from Baku State University. Her current research interests include the automatic generation of assertion oracles, error propagation in software programs, mutation testing of deep learning systems, testing of ethics in AI systems and oracles for autonomous vehicles.