- Shahroz Tariq, Binh M Le, and Simon S Woo. “Towards an Awareness of Time Series Anomaly Detection Models’ Adversarial Vulnerability” at CIKM ’23 (CORE A). This work demonstrates that the performance of state-of-the-art DNN- and GNN-based anomaly detection methods are degraded substantially by adding only small adversarial perturbations to the sensor data. The overarching goal of this research is to raise awareness of the adversarial vulnerabilities of time series anomaly detectors.
- Chaoran Li (Swinburne University), Xiao Chen (Monash University), Ruoxi Sun (CSIRO’s Data61), Jason Xue (CSIRO’s Data61), Sheng Wen (Swinburne University), Muhammad Ejaz Ahmed (CSIRO’s Data61), Seyit Camtepe (CSIRO’s Data61), and Yang Xiang (Swinburne University), “Cross-Language Android Permission Specification”, ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE), 2022
- Chunyi Zhou (Nanjing University of Science and Technology), Yansong Gao (Nanjing University of Science and Technology), Anmin Fu (Nanjing University of Science and Technology), Kai Chen (Chinese Academy of Science), Zhiyang Dai (Nanjing University of Science and Technology), Zhi Zhang (CSIRO’s Data61), Jason Xue (CSIRO’s Data61), Yuqing Zhang (Chinese Academy of Science), PPA: Preference Profiling Attack Against Federated Learning, The Network and Distributed System Security Symposium (NDSS), 2023
- Two papers were just accepted. CSIRO leads NeurIPS paper and significantly contributes to NDSS 2023 in terms of the core idea and experiments and overall writing.
[NDSS 2023] Wanlun Ma (Swinburne University of Technology), Derui Wang (CSIRO’s Data61), Ruoxi Sun (The University of Adelaide & CSIRO’s Data61), Minhui Xue (CSIRO’s Data61), Sheng Wen (Swinburne University of Technology), Yang Xiang (Digital Research & Innovation Capability Platform, Swinburne University of Technology), The “Beatrix” Resurrections: Robust Backdoor Detection via Gram Matrices, The Network and Distributed System Security (NDSS), 2023
[NeurIPS 2022] Pingyi Hu (University of Adelaide), Zihan Wang (University of Adelaide), Ruoxi Sun (The University of Adelaide & CSIRO’s Data61), Hu Wang (University of Adelaide), Minhui Xue (CSIRO’s Data61), M^4I: Multi-modal Models Membership Inference, Conference on Neural Information Processing Systems (NeurIPS), 2022
- Chandra Thapa, Seyit Camtepe, Raj Gaire, Surya Nepal, Seung Ick Jang, “Demo – MaLFraDA: A Machine Learning Framework with Data Airlock,” accepted and to be presented at ACM CCS’22 (Core A*). This work proposes a new multi-zoned framework called MaLFraDa for training machine learning algorithms on sensitive, illegal to possess and psychologically harmful data. MaLFraDA has soft air gaps between its zones to isolate and control communication in and out of the framework. With an extension, the framework runs popular distributed machine learning algorithms such as federated and split learning using multiple data custodians.
- One paper accepted: IEEE TCAD 2022: Title: RBNN: Memory-Efficient Reconfigurable Deep Binary Neural Network with IP Protection for Internet of Things. Authors: Huming Qiu (Nanjing University of Science and Technology), Hua Ma (The University of Adelaide and CSIRO’s Data61), Zhi Zhang (CSIRO’s Data61), Yansong Gao (Nanjing University of Science and Technology), Yifeng Zheng (Harbin Institute of Technology), Anmin Fu (Nanjing University of Science and Technology), Pan Zhou (Huazhong University of Science and Technology), Derek Abbott (The University of Adelaide), and Said F. Al-Sarawi (The University of Adelaide)
Erik Buchholz (UNSW Sydney), Sharif Abuadbba, Shuo Wang, Surya Nepal; Salil S. Kanhere, “Reconstruction Attack on Differential Private Trajectory Protection Mechanisms” accepted at the DSS target security venues Annual Computer Security Applications Conference (ACSAC) 2022.
- Zhu, Fei, Xun Yi, Sharif Abuadbba, Ibrahim Khalil, Surya Nepal, and Xinyi Huang. “Authenticated Data Sharing with Privacy Protection and Batch Verification for Healthcare IoT.” got paper accepted at IEEE Transactions on Sustainable Computing (2022).
Authors: Cody Christopher, Kristen Moore, David Liebowitz (external, CSCRC)
Paper: “SchemaDB: A Dataset for Structures in Relational Data”
Description: We have constructed and are releasing a new dataset to enable the study of relational structures in structured databases. This was born from a need to be able to generate artifacts to use for cyber deception purposes, which requires a training set of examples to learn from. We present this dataset, as well as outline some potential research avenues and broader applications.
Conference: AusDM’22 THE 20TH AUSTRALASIAN DATA MINING CONFERENCE 2022 – Western Sydney, 12 -15 Dec
- Falih Gozi Febrinanto, Feng Xia, Kristen Moore, Chandra Thapa, and Charu Aggarwal, “Graph Lifelong Learning: A Survey”, accepted at IEEE Computational Intelligence Magazine (IF: 11.356). This paper reviews graph learning techniques that can learn incrementally over new tasks on evolving graphs and maintain performance from previous tasks from catastrophic forgetting. This survey paper aims to open different perspectives and opportunities in solving dynamic graph-based problems with a continual and efficient learning process.
- Chandra Thapa, Seung Ick Jang, Muhammad Ejaz Ahmed, Seyit Camtepe, Josef Pieprzyk, Surya Nepal, “Transformer-Based Language Models for Software Vulnerability Detection,” accepted at The Annual Computer Security Applications Conference (ACSAC) [DSS target list conference]. This work studies how to leverage (large) transformer-based language models in detecting software vulnerabilities and how good these models are for vulnerability detection tasks. In this regard, firstly, we present a systematic (cohesive) framework that details source code translation, model preparation, and inference. Then, we perform an empirical analysis of software vulnerability datasets. Our results demonstrate the good performance of the models. Moreover, this paper also analyses the popular platforms to efficiently fine-tune these models and present recommendations while choosing the platforms for our framework.
- Bushra Sabir, Muhammad Ali Babar, Raj Gaire, Alsharif Abuadbba, “Reliability and Robustness analysis of Machine Learning based Phishing (MLPU) URL Detectors”, accepted at IEEE Transactions on Dependable and Secure Computing 2022 [DSS target publication venue]. This work carries out the most extensive systematic study to investigate 50 existing Machine learning-based phishing URL detectors. The work firstly introduced a cost-effective adversarial URL generator named URLBUG. The work findings identified several security vulnerabilities of these systems and provided future direction to design secure MLPU systems.
- Mohan Baruwal Chhetri and Marthie Grobler hosted the third Human Centric Software Engineering & Cyber Security workshop with IEEE/ACM Automated Software Engineering conference on 11 October.
- Muhammed Esgin is part of an application for a US Department of State grant on post-quantum cryptography training and capability assessment that has been successful.
- Hajime Suzuki participated in ITU-R Working Party 5D meeting from 10 – 21 October 2022 (remote participation) as part of Department of Home Affairs 6G Future Connectivity Research Partnership (OD-228762) project, following the development of 6G standards. Preliminary draft Recommendation on 6G vision and requirements is expected to be released in February 2023.
- Professor Josef Pieprzyk awarded IACR Fellowship, first Australian in the last five years to receive this recognition.
Professor Josef Pieprzyk of CSIRO’s Data61 can now add International Association for Cryptologic Research (IACR) Fellow to his list of outstanding achievements in the field of data security. Below, Josef looks back on his career milestones, most impactful research, and proudest moments, and shares some insights into his next big breakthrough. https://algorithm.data61.csiro.au/professor-josef-pieprzyk-awarded-iacr-fellowship/
- The PIF project (Automatic Assessment and Protection of Personal Information for Data Sharing) was one of the finalists in the iAwards national event under the Technology Platform Solution Award of the Year. The national iAwards ceremony was held on the 20th of October, 2022, at Q-Events by Metropolis, Melbourne. Although PIF couldn’t win an award, this national recognition was an outstanding achievement for the project.
- Smartshield made it to the iAwards national finals ceremony in Melbourne after winning the technology platform category in NSW mid-July 2022. Smartshield is using AI to combat phishing in an explainable way to help users and security analysts to understand the decision-making process. While we did not win, but it was amazing experience to compete against the finest nationally across the states.https://research.csiro.au/cybersecurity-quantum-systems/artificial/ / https://cybersecuritycrc.org.au/smart-shield-artificial-intelligence-anti-phishing-system
In the last 6 weeks DSS group has welcomed: 5 students: Benoit, Guillaume, Eliot, Pierre and Haodong; 5 postdocs: Ruoxi Sun and Hongsheng Hu, Autumn Wu, Raymond Zhao and Akib Karim, 1 visiting Scientist: Zihan Wang. Welcome to all!
- Meet Hongsheng:
Hello everyone! I am Hongsheng Hu. I did my Ph.D. at The University of Auckland in New Zealand. Now, I have moved to Sydney and started my post-doctoral job at CSIRO. I am really excited to join the DSS group! My research focuses on AI privacy and security, especially membership inference attacks and distributed systems security. If you are working on such topics and want to have a discussion and collaboration, feel free to contact me via any means.
The new Quantum team welcome 3 postdocs:
- Autumn W, FSP project: Autumn Wu, a new postdoctoral fellow for Quantum ML for Cyber project under Quantum Technologies FSP has commenced on the 31st of October 2022. Autumn conducted her PhD research in off-policy and off-line deep reinforcement learning at the New York University.
- Kai (Raymond) Zhao -> IWOY project: I am a new postdoctoral fellow who just started with the CSIRO’s Data61 at the beginning of November. Before joining CSIRO’s Data61 I received my PhD and worked as a research fellow in the Department of Software Systems and Cybersecurity, Faculty of Information Technology, Monash University, Australia, in 2022. My research focuses on efficient and secure implementation techniques for post-quantum cryptographic applications and protocols. I feel quite excited to join such an innovative and diverse research team in CSIRO’s Data61. I hope my research at CSIRO’s Data61 can deliver visible impact on the transitioning into a quantum-safe world by improving the implementation of post-quantum cryptography (PQC) and helping the integration of PQC.
- Akib Karim -> IWOY project: Akib is a postdoctoral fellow at CSIRO’s Data61. He has a PhD in Quantum Chemistry from RMIT University and BSc with Honours in Physics from University of Sydney. He has research interests in Quantum Information Theory, Quantum Algorithms, and Quantum Chemistry.
Visit our career pages:
CSIRO Industry PhD Scholarship: AI techniques in intelligent manufacturing
Project Title: Data-driven robust, explainable AI techniques for process, product quality control, and security in intelligent manufacturing
Project Description: Quality, productivity, and security are essential elements in an industrial production plant. Even a slight improvement in productivity, e.g., 1%, can lead to gains of Millions in revenue. These elements are dependent on various factors, including process control and automation. Furthermore, these factors are derived from several probabilistic and deterministic parameters that span from raw material collection and transportation to manufacturing. In this 4-year Ph.D. research project, a student will investigate the novel methods to improve the quality (e.g., customer requirements), productivity (e.g., production time), and security of manufacturing by considering data-driven approaches and leveraging robust explainable artificial intelligence/machine learning algorithms for overall intelligent automation. This project is outcomes-oriented, so the data is collected from the actual plant of Sonac Australia Pty Ltd, an animal feed industry, and the proposed control methods and techniques will be leveraged to optimize the actual plant.
The project will focus on the process of blood collection and delivery to the production plant. Despite its importance for the quality of the process outcome, this part of the production cycle does not allow for best control and monitoring as it takes place at slaughterhouses and on delivery tankers. The project will look at ways of automating the initial blood processing in the slaughterhouse in a way which prevents damage to the blood (e.g., premature coagulation, impurities, etc.). Also, the process of pumping blood to and from the transporting tanks will be a target for improvement as it has settings (e.g., pumping flowrate) which impact the quality and productivity of the Sonac processes. The outcome of this project will result in better controllability over the pre-plant processing.
CSIRO supervisor: Chandra Thapa & Seyit Camtepe
Email address: Chandra.firstname.lastname@example.org; and Seyit.Camtepe@data61.csiro.au
University Federation University
Our group is welcoming 4 french students. Let’s hear from them:
- Bonjour le CSIRO ! Student in a French engineering school, I have taken a gap year to discover other fields, such as research, and also travel and improve my communication skills. Australia is perfect for this ! English environment, exotic country, and the opportunity to work with CSIRO could not be refused ! Clayton site is very modern, feels very good from the beginning, and everybody is so welcoming, it I a pleasure to see how nice the people are here. I work with Benoit Morvan, another French student, under the supervision of Tina Wu. The project I am taking part in aims to create an interactive dashboard allowing the user to visualize the personal information that are collected. This dashboard must include the data collected by a VR headset in order to highlight the new data collected and therefore at risk when using such devices. I am very excited to work in here, and I can’t wait to see where this project will take me. See you on Clayton campus ! Guillaume
- We had the opportunity to spend 3 months at Data6! Working in cyber security. The research topic allowed us to go outside our skill base but also the discovery of Australian research. The CSIRO regularly organizes interventions on a wide variety of research topics, allowing us to constantly discover new technological advances. The different activities proposed by the researchers (Morning Tea or Barbecue) allow us to discuss and learn more about the work of the researchers present ! It is therefore a rich experience from a personal point of view (with a foretaste of tomorrow’s technologies) but also professionally. Benoit
News from our students:
- Jodi Gagner (from Griffith University) is working as an undergraduate trainee student (from 18/07/2022 to 18/10/2022) on “Privacy-Preserving Biometric Recognition” with M.A.P. Chamikara.
- 1st quantum student, Ben Harper, started end of October with Usman Muhammad.
- Ms Farina Riaz, 2nd year R+ top-up scholarship PhD student with University of Southern Queensland (Prof Ravinesh Deo and Dr Shahab Abdulla) joined group. She is currently working on the application of Quantum ML for traffic sign recognition. Riaz is participating in the 3MT competition for UniSQ and her abstract was accepted in the conference Quantum Machine Learning and Applications happening in Sydney in Dec 2022. Presentations from UniSQ College included Ph.D. researcher, Farina Riyaz (who is supervised by a team including Shahab Abdulla, Ravinesh Deo and Susan Hopkins. The topic of Farina’s Three Minute Thesis was “Next Generation Image Classification for Intelligent Transportation Systems utilising Quantum Machine Learning”
- Farina Riaz is participating in her university 3 Minutes Thesis competition.
CSIRO has joined the Australian Quantum Software Network along with universities and start-ups. Dr Usman, Team Leader Quantum Systems, will be representing Data61’s Quantum Software Research. The aim of this group of experts is to help keep Australia at the forefront of quantum software technologies. Further details are here: https://www.innovationaus.com/quantum-software-alliance-forms-to-keep-australia-ahead/
- Our Quantum team leader Usman Muhammad, participated at the exhibition of Quantum Adversarial Machine Learning Technology at the 2022 Chief of Army Symposium in Adelaide from August 10-11.
- Usman Muhammad did a Guest Lecture at the Australian Institute of Machine Learning (AIML), The University of Adelaide on August 12, 2022, Applications at the intersection of machine learning and quantum computing.
- Zhi Zhang has been invited to serve as a reviewer for ACM Computing Surveys 2022 (CORE A*) and IEEE Transactions on Dependable and Secure Computing 2022 (CORE A).
- Mohan Baruwal Chhetri and Shahroz Tariq presented at the CINTEL FSP Meet and Mingle in Canberra on 24 and 25 October.
- Kristen Moore was part of a Q and A session panel where three Cybersecurity Professionals share their experiences within the industry and discuss all things cybersecurity. This was part of CyberTaipan.
- Ejaz Ahmed, Sharif Abuadbba, Cody Christopher, and Surya Nepal participated in a full-day workshop for our DSTG malware triaging project with our partners, including DSTG, Macquarie University, University of Newcastle – in Newcastle on 18 October 2022.
- Sharif Abuadbba, Shuo Wang, M.A.P Chamikara, Surya Nepal attended the iAwards national finals ceremony to represent our 2 projects that won the iAwards Technology platforms in NSW – in Melbourne 20 October 2022.
- Shahroz Tariq arranged a collaboration between Sungkyunkwan University (SKKU) from South Korea and Data61 (Sharif and Kristen) on deepfakes and security.
- Helge Janicke, Research Director Cyber Security CRC visited our Sydney and Adelaide office.
SAO seminars in collaboration with the Cyber Security CRC: https://research.csiro.au/cybersecurity-quantum-systems/our-sao-seminars/
Human Centric seminars: https://research.csiro.au/cybersecurity-quantum-systems/the-human-centric-ai-seminars-series/
DSS group is launching Quantum Systems monthly seminar series, for more info
Data61 has established a new quantum technology program, focused in the areas of quantum software, quantum security, and quantum algorithms & applications. This seminar series will invite quantum experts to provide an updated summary of the global research on the topics of interest, highlight key challenges in the development of quantum technologies and stimulate new ideas for future research directions. The seminar series will also provide engagement and networking opportunities for Data61 researchers. The seminars will be scheduled on the last Wednesday (3-4 PM AEST) of every month.
In collaboration with Quantum Technology FSP (Prof Jim Rabeau); for more info https://research.csiro.au/qt/