Ajitha Rajan

arajan AT ed.ac.uk

My Research

• Automated Software Testing Techniques considering test input generation, test oracles, coverage metrics.
• Biomedical Artificial Intelligence focusing on cancer survival models, interpretability for biological sequences and medical images. .

I am looking for PhD students with strong background in either

• Software testing, program analysis and compilers. Students should have very strong programming skills, preferably in C++ and Python.
• Natural language processing or computer vision with some familiarity in bioinformatics. Students should have solid understanding of AI techniques and tools along with strong Python programming.

Teaching

• Software Testing (ST) - 3rd year UG course introducing software testing concepts and providing practical experience with JUnit. I have been teaching this course since 2015.
• Computer Programming - Skills and Concepts (CP) - 1st year UG course introducing C programming. I co-taught this course in 2017/18 and 2018/19.
• Embedded Systems (ES) - 4th year UG course (2017/18).
• Introduction to Software Engineering (Inf2C-SE) - 2nd year UG course (2014/15).

Recent News

2023

• December: Our paper with collaborators at Bytedance Research with Chao Peng as lead author, "Hawkeye: Change-targeted Testing for Android Apps based on Deep Reinforcement Learning", was accepted at ICSE 2024, Software Engineering In Practice track.
• December: Our paper with Xiaoliang Wu as lead author, "Can We Trust Explainable AI Methods on ASR? An Evaluation on Phoneme Recognition", was accepted at ICASSP 2024.
• August: Our paper with Nick Louloudakis as lead author, "DeltaNN: Assessing the Impact of Computational Environment Parameters on the Performance of Image Recognition Models", was accepted at ICSME 2023, Industry track.
• July: Our paper with Nick Louloudakis as lead author, "Fault Localization for Buggy Deep Learning Framework Conversions in Image Recognition", was accepted at ASE 2023, NIER track.
• June: Our paper with Benedicte Legastelois as lead author, "Challenges in Explaining Brain Tumor Detection", was accepted at TAS 2023.
• March: Our paper with Georges Bedran as lead author, "The immunopeptidome from a genomic perspective: Establishing the non-canonical landscape of class I MHC-associated peptides." was accepted in the Cancer Immunology Research journal3. To Appear.
• February: Gave an invited talk at Royal Society Meeting of Minds Conference on "Ensuring Safety and Robustness of Perception AI in Autonomous Vehicles" in the Data and Digital Technology Theme.
• February: Our paper with Xiaoliang Wu as lead author, "Explanations for Automatic Speech Recognition" was accepted at IEEE ICASSP 2023. To Appear in June 2023.

  2022

• November: Our paper with Nick Louloudakis as lead author,"Assessing RObustness of Image Recognition Models to Changes in the Computational Environment" was accepted and presented at NeurIPS ML Safety Workshop 2022.
• September: Our paper with Amy Rafferty as lead author,"Explainable Artificial Intelligence for Breast Tumour Classification: Helpful or Harmful" was accepted and presented at Workshop on Interpretability of Machine Intelligence in Medical Image Computing at MICCAI 2022.
• August: Our paper with Xiaoliang Wu as lead author, "Catch Me If You Can: Blackbox Adversarial Attacks on Automatic Speech Recognition using Frequency Masking" was accepted at Proceedings of 2022 29th Asia-Pacific Software Engineering Conference (APSEC). To Appear in Dec 2022.
• August: Our paper with Foivos Tsimpourlas as lead author," BenchPress: A Deep Active Benchmark Generator" was accepted at 31st International Conference on Parallel Architectures and Compilation Techniques (PACT). To Appear October 2022
• June: Our paper with Dominic Phillips, Hans-Christof Gasser, Sebestyén Kamp as lead authors,"Generating Immune-aware SARS-CoV-2 Spike Proteins for Universal Vaccine Design" was accepted in Proceedings of the 1st Workshop on Healthcare AI and COVID-19, ICML 2022, PMLR 184:100-116, 2022. July 2022.
• May: PC Member for ICSE 2023 (SEIP track) and 2024,FSE 2023, ISSTA 2023.
• April: Invited talks at Stanford University Dept. Aeronautics and Astronautics Seminar and Facebook AI Research in California.

  2021

• October: Our paper,"Interpreting BERT architecture predictions for peptide presentation by MHC class I proteins", is accepted at NeurIPS Workshop 2021, "Explainable AI approaches for Debugging and Diagnosis".
• October: Awarded Royal Society Industry Fellowship on "AutoTest: Testing Autonomous Vehicle Perception Safety on Hardware Accelerators". The fellowship will start March 2022.
• August: Our paper,"GPU Acceleration of FSM Input Execution: Improving Scale and Performance", is accepted at STVR journal 2021.
• June: Our paper, "Testing Smart Contracts: Which Technique Performs Best?", is accepted at ACM/IEE Conference ESEM 2021.
• June: Our paper, "CAT: Change-focused Android GUI Testing", is accepted at IEEE conference ICSME 2021.
• June: Our paper, "Embedding and Classifying Test Execution Traces using Neural Network", is accepted in IET Software Journal.
• May: Our paper, "GPU Acceleration of FSM Input Execution: Improving Scale and Performance", is accepted with minor revisions in STVR Journal.
• May: PC Member for FSE 2021 Industry track, FASE 21, ISSTA 21 Posters Chair, ICSE 2022 SEIP track, FSE 2022 Research Paper and Industry Track, ASE 2022 Journal First Track.

I edited a book along with Thomas Wahl for the European funded project, CESAR. The book was published by Springer in March 2013. A glimpse of the book is available here.