News Feed January February 2023

May 24th, 2023

Published papers

  • Haonan Zhong, Jiamin Chang, Ziyue Yang, Tingmin Wu, Pathum Chamikara Mahawaga Arachchige, Chehara Pathmabandu, and Minhui Xue. 2023. Copyright Protection and Accountability of Generative AI: Attack, Watermarking and Attribution. In Companion Proceedings of the ACM Web Conference 2023 (WWW ’23 Companion), April 30-May 4, 2023, Austin, TX, USA. ACM, New York, NY, USA, 5 pages. https://doi.org/10.1145/3543873.3587321. This paper proposes an evaluation framework to provide a comprehensive overview of the current state of the copyright protection measures for GANs, evaluate their performance across a diverse range of GAN architectures, and identify the factors that affect their performance and future research directions.
  • Farina Riaz (CSIRO’s Data61, University of Southern Queensland), Shahab Abdulla (UniSQ), Hajime Suzuki (Data61), Srinjoy Ganguly (UniSQ), Ravinesh C. Deo (UniSQ), and Susan Hopkins (UniSQ), “Accurate image multi-class classification neural network model with quantum entanglement approach,” Sensors (Impact Factor 3.85), March 2023. 11 pages. https://doi.org/10.3390/s23052753. This is the first paper published from the new Quantum ML for Cyber project. In this paper, we propose a novel neural network with quantum entanglement method to enhance the accuracy of image classification. We show the improvements in image classification accuracy over 10 class hand written digits and photos but not over 43 class real-life traffic signs. Exact causes of the improvement and degradation are currently open questions and requiring further investigation.

Publications accepted

  • CVPR 2023, C. Chen, D. Liu, S. Ma, S. Nepal, C. Xu: Private Image Generation with Dual-Purpose Auxiliary Classifier, Proceedings of The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023.
  • Venue: The Web Conference (WWW) 2023; Title:AgrEvader: Poisoning Membership Inference against Byzantine-robust Federated Learning. Authors: Yanjun Zhang, Guangdong Bai, Pathum Chamikara Mahawaga Arachchige, Mengyao Ma, Liyue Shen, Jingwei Wang, Surya Nepal, Minhui Xue, Long Wang and Joseph K. Liu. This paper proposes a poisoning membership inference attack that maximizes the adversarial impact to circumvent the attack detection by Byzantine-robust mechanisms.
  • Venue: IEEE S&P 2023; Title: BlindHub: Bitcoin-Compatible Privacy-Preserving Payment Channel Hubs Supporting Variable Amounts; Authors: Xianrui Qin (The University of Hong Kong), Shimin Pan (The University of Hong Kong), Arash Mirzaei (Monash University), Zhimei Sui (Monash University), Oğuzhan Ersoy (Radboud University and Delft University of Technology), Amin Sakzad (Monash University), Jiangshan Yu (Monash University), Muhammed F. Esgin (Monash University and CSIRO’s Data61), Joseph K. Liu (Monash University), Tsz Hon Yuen (The University of Hong Kong)
  • Six papers accepted to the top security conferences, IEEE Security and Privacy (Oakland) and NDSS. The conferences will be held next May in San Francisco and next March in San Diego, respectively.

[Oakland 2023]; Title: BlindHub: Bitcoin-Compatible Privacy-Preserving Payment Channel Hubs Supporting Variable Amounts; Authors: Xianrui Qin (The University of Hong Kong), Shimin Pan (The University of Hong Kong), Arash Mirzaei (Monash University), Zhimei Sui (Monash University), Oğuzhan Ersoy (Radboud University and Delft University of Technology), Amin Sakzad (Monash University), Jiangshan Yu (Monash University), Muhammed F. Esgin (Monash University and CSIRO’s Data61), Joseph K. Liu (Monash University), Tsz Hon Yuen (The University of Hong Kong)

[Oakland 2023] Shuo Wang (Data61), Sharif Abuadbba, Sidharth Agarwal, Kristen Moore, Ruoxi Sun, Minhui Xue, Surya Nepal, Seyit Camtepe, and Salil Kanhere, PublicCheck: Public Integrity Verification for Services of Run-time Deep Models, IEEE Symposium on Security and Privacy (Oakland), 2023

The collaboration was done within the DSS team. This is related to the mission for Responsible AI.

[Oakland 2023] Yuxin Cao, Xi Xiao, Ruoxi Sun, Derui Wang, Minhui Xue, and Sheng Wen, StyleFool: Fooling Video Classification Systems via Style Transfer, IEEE Symposium on Security and Privacy (Oakland), 2023

The collaboration was done with Swinburne University and Tsinghua University. We provided a state-of-the-art adversarial attack on video classifiers, resonating with multimedia misinformation during warfare. With Derui and Ruoxi from Data61, our role is to help students formalize the idea, develop the methodology and hands on writing. This is related to the mission for Responsible AI.

[NDSS 2023] Tian Dong, Shaofeng Li, Guoxing Chen, Minhui Xue, Haojin Zhu, and Zhen Liu, RAI^2: Responsible Identity Audit Governing the Artificial Intelligence, The Network and Distributed System Security (NDSS), 2023

The collaboration was done with Shanghai Jiao Tong University. We provided a responsible AI copyright check system. My role is to supervise two ongoing PhD students at Shanghai Jiao Tong University. I hope to get the first student, Dong Tian, to come to Australia for research internship for a year. This is related to the mission for Responsible AI.

[NDSS 2023] Wanlun Ma, Derui Wang, Ruoxi Sun, Minhui Xue, Sheng Wen, and Yang Xiang, The “Beatrix” Resurrections: Robust Backdoor Detection via Gram Matrices, The Network and Distributed System Security (NDSS), 2023

The collaboration was done with Swinburne University. We provided a backdoor attack detection using Gram Matrices. With Derui and Ruoxi from Data61, our role is to help students formalize the idea, develop the methodology and hands-on writing. This is related to the mission for Responsible AI.

[NDSS 2023] Shuo Wang (Data61), Mahathir Almashor, and Sharif Abuadbba; Ruoxi Sun, Minhui Xue, and Calvin Wang, Raj Gaire, Seyit Camtepe, and Surya Nepal, DOITRUST: Dissecting On-chain Compromised Internet Domains via Graph Learning

The work is an output of our Smartshield CRC project. It is a domain intelligence trust mechanism that leverage graph learning to learn and predict not only the malicious domains, but also the benign but compromised domains that are exploited by the attackers to launch cyber attacks.

  • Shahzeb Tariq (KHU Korea), Shahroz Tariq (CSIRO’s Data61), SangYoun Kim (KHU Korea), Simon S. Woo (SKKU Korea), Chang Kyoo Yoo (KHU Korea). “Distance adaptive graph convolutional gated network-based smart air quality monitoring and health risk prediction in sensor-devoid urban areas” accepted at the Sustainable Cities and Society Journal [IF: 10.696]. In this work, we predict the air pollution levels at sensor-devoid locations. We propose a distance-adaptive graph convolutional gated network that provides simultaneous forecasts of primary pollutants at multiple megacity locations and temporal horizons by fusing spatiotemporal sensor data.
  • Shahroz Tariq (CSIRO’s Data61), Sowon Jeon (SKKU Korea), Simon S. Woo (SKKU Korea), “Evaluating Trustworthiness and Racial Bias in Face Recognition APIs using Deepfakes” accepted at the IEEE Computer Magazine [IF: 3.564 (2018)]. This work evaluates the racial bias in face recognition APIs using real and deepfake celebrity images. We show how deepfake images exacerbated racial bias in Amazon, Microsoft, and Naver web-based face recognition APIs. The findings are significant because they reveal similar vulnerabilities to those previously discovered through adversarial attacks but through a significantly different method.
  • Baiqi Chen, Tingmin Wu, Yanjun Zhang (Deakin University), Mohan Baruwal Chhetri, Guangdong Bai (University of Queensland). Investigating Users’ Understanding of Privacy Policies of Virtual Personal Assistant Applications. Accepted at AsiaCCS 2023. The paper presents the results of a subjective study to investigate the level of users’ understanding of privacy policies, targeting VPA apps of Amazon skills. The study focussed on users’ understanding of technical terms in privacy policies and the role of explanations in improving users’ understanding.
  • Mengyao Ma, Yanjun Zhang (Deakin University), Leo Yu Zhang (Deakin University), M.A.P. Chamikara, Mohan Baruwal Chhetri, Guangdong Bai (University of Queensland). LoDen: Making Every Client in Federated Learning a Defender Against the Poisoning Membership Inference Attacks. Accepted at AsiaCCS 2023. The paper proposes a novel client-side defence mechanism against poisoning membership inference attacks in federated learning by leveraging access to own datasets to detect suspicious privacy attacks against samples and remove the ones under attack.
    Geetanjli Sharma, M.A.P. Chamikara, Mohan Baruwal Chhetri, Yi-Ping Phoebe Chen (La Trobe University). SoK: Systematising Attack Studies in Federated Learning – From Sparseness to Completeness. Accepted at AsiaCCS 2023. The paper uses a causal model inspired by structural equation modelling to systematise the existing literature on FL attack studies.

Projects

6G Security and Privacy Seminars kicked-off with the vice president of Samsung Research India providing the first presentation, aiming to grow Australian cybersecurity capacity and capability, and be ready with a resilient roadmap for 6G.

Talks and Conferences

Our team participated in a three day Alan Turing workshop, in Eveleigh 13/14/15 March 2023.

The Alan Turing Institute and CSIRO Data61 Workshop was a highly productive event that took place in Sydney from March 13th to March 15th, 2023. The purpose of the workshop was to explore strategic collaborations in the areas of cybersecurity and sustainability. The participants from our group: Surya Nepal, Sharif Abuadbba, Seyit Camtepe, Ejaz Ahmed, Jason Xue, Mohan Barruwl Chhetri, Chandra Thapa.

During the event, both CSIRO’s Data61 and Alan Turing presented their program capabilities and new initiatives. Subsequently, the teams collaborated to brainstorm ideas and identify key areas of focus for the next three years. This included discussing policies and the potential impact of proposed projects, identifying scientific gaps, and outlining areas of focus. As a result of this collaboration, the teams were able to develop a number of concrete ideas that they plan to pursue in the future.

The workshop concluded with discussions regarding future collaborations, including potential areas of focus for the ATI-CSIRO Symposium, the development of a governance model, funding plans, and ways to include third-party organisations and universities in these collaborative efforts. Overall, the workshop was a success and laid the foundation for continued collaboration between these two organisations in the future.

  • This year Sydney Quantum Academy were proud to showcase Australia’s wealth of research talent with over 50 innovative research posters hosted at the Quantum Australia 2023 in-person event. Our Postgraduate PhD Student Farina Riaz research received award runners-up for the $250 People’s Choice Award. Congratulations to Farina Riaz (CSIRO/ University of Southern Queensland) for the poster titled Performance Comparison of Quantum Machine Learning Models – Image Classification for Future AI Applications.
  • Cécile Paris, Surya Nepal, Mohan Baruwal Chhetri, Fatemeh Jalalvand are organising as session in the Modelling and Simulation (MODSIM) 2023 conference next year. Please consider submitting an extended abstract or full paper to the Session on Simulation-based Analysis for Cybersecurity Decision-Making at MODSIM2023
    contact: Cécile Paris, Surya Nepal, Mohan Baruwal Chhetri, Fatemeh Jalalvand
  • Hajime Suzuki gave a 10 minutes presentation on Department of Home Affairs 6G Security Research and Development Project to Digital Technical Standards Points of Contact Group consisting of government officers from UK, Europe, USA, etc. on 1 December 2022, promoting international collaboration.
  • Usman Muhammad was on a panel at the 2022 South Australia Investment Conference in Adelaide. Topic: South Australia’s future in developing opportunities in the global applications of quantum technologies. Further information here: https://www.saic2022.au/
  • Our group enjoyed the visit of Prof. Lam from NTU, Singapore, 23 Nov 2022 at our Marsfield site. He also presented his research during one of our CRC/Data61 SAO seminar. Title: Digitalization, Digital Trust and TrustTech
  • Sydney team including Sharif Abudbba, Chamikara Mahawaga Arachchige attended TCS APAC summit in Sydney on the 8-9 November 2022 as part of cybersecurity crc to represent few of our projects like Smartshield, PIF, Gamifications, etc. joined by Helge Janicke, Cyber Security CRC.
  • Josef Piperzyk is serving as Editor for Topical Collection “Survey of Cryptographic Topics” — MDPI Cryptography https://www.mdpi.com/journal/cryptography/topical_collections/6LY07YAMY2; Topic: Key Management and Key Recovery – Special Issue of Journal of Surveillance, Security and Safety (jointly with Moti Yung) https://jsssjournal.com/journal/special_detail/1346
  • Mohan Baruwal Chhetri is on the SERVICES Congress Advisory Program Committee for SERVICES 2023
  • Marthie Grobler presented the Critical Infrastructure Protection and Resilience mission at the Digital Twins, Simulation & Modelling CCC Community of Practice on 2 February 2023, Webinars – Digital Twins, Simulation & Modelling CCC Community of Practice – Confluence (csiro.au).
  • “When humans are involved, cyber becomes complicated”. Listen to Dr Marthie Grobler and others share their thoughts on phishing, online scams and human centric security. ‘Anatomy of a scam — episode ‘Gone Phishing: the racehorse trainer, the Mercedes-Benz, and the missing $100,000’ follows the journey of scam emails and calls, and why the human factor is so critically important in educating users to become more resilient against phishing attacks. https://www.nine.com.au/entertainment/podcasts/anatomy-of-a-scam?utm_source=D61SocialMedia&utm_medium=D61SocialMedia&utm_campaign=AoaS&utm_content=AoaS&utm_medium=newsletter&utm_campaign=InsideData61-InsideData61-Edition-60&utm_source=InsideData61-InsideData61-Edition-60-email-20221130
  • Our very own Dr Kristen Moore attended Science Meets Parliament in Canberra. for more information https://scienceandtechnologyaustralia.org.au/what-we-do/science-meets-parliament/

  • Dr Usman was invited as a panel member on Quantum Careers: Academia vs Industry panel at the Quantum Australia 2023 conference. The panel compared career choices in academia and industry for students and ECRs working in the field of quantum computing.

Staff

Welcome to Bahar Goldozian
who just joined our newly formed Quantum Systems team

Bahar holds a PhD in Theoretical Physics from Lund University, where she specialized in electron transport in nano structures. Her work was centred around open quantum systems that are out of equilibrium. These systems included quantum coherent electronic conductors that were influenced by temperature and/or voltage changes, along with other platforms that hold potential for emerging quantum technologies. She utilized principles from quantum mechanics, thermodynamics, and electrodynamics to understand devices where quantum phenomena play a significant role. As an expert in quantum mechanics, she deeply understands this complex field and has contributed to developing innovative software tools to solve quantum mechanics problems. Currently, Bahar is a postdoctoral researcher in the field of quantum computing, where they are working to advance our understanding of this cutting-edge technology and its potential applications.

Welcome to Behnam
who just joined our newly formed Quantum Systems team

I am a theoretical quantum physicist. I did my PhD at University of Queensland within ARC centre for Engineered Quantum Systems (EQUS). During my PhD, I designed quantum devices such as an autonomous quantum heat engine and protocols for high precision measurement exploiting quantum coherence in cold atoms. Following my PhD, I worked as a postdoctoral fellow at Griffith University and the ARC Centre for Quantum Computation and Communication Technology (CQC2T) on protocols for optimum control and noise suppression in quantum devices. During my time at CQC2T, I became interested in quantum algorithms, particularly hybrid quantum classical algorithms. Now in my role at CSIRO, I am working on hybrid quantum-classical algorithm to achieve higher computational power of quantum devices. Also I am working on quantum control: how to control quantum system faster and more precise.

Welcome to Ronal Singh

Ronal is a Research Scientist with the Human Centric Security team in CSIRO’s Data61 and a member of the CINTEL FSP. Ronal’s primary interest lies in artificial intelligence, particularly multi-modal human-agent interactions, AI-assisted decision-making, explainable AI, intention recognition, and multiagent communication planning. Previously, he was a Research Fellow in Human-Agent Collaboration and an Associate Lecturer in the School of Computing and Information Systems at the University of Melbourne. Ronal completed his PhD in 2018 from the University of Melbourne and his BSc and MSc degrees in Computer Science from the University of the South Pacific in the Fiji Islands.

Welcome to Sarah Siddiqui

Sarah joined CSIRO’s Data61 as a CERC Postdoctoral Fellow in January 2023. Before joining CSIRO, Sarah completed her PhD from Macquarie University Australia where she worked on security of the Internet of vehicles. She is particularly interested in the intelligent transportation systems, trust management, and AI/ML-based security.

Welcome to Jiafan Wang

Jiafan Wang is a postdoctoral research fellow at Data61. He received his Ph.D. degree from the Department of Information Engineering, the Chinese University of Hong Kong, in 2022. He has published several papers in the area of applied cryptography and served as a reviewer for reputable conferences and journals. His current research interest includes encrypted data management and post-quantum cryptography.

Students Update

  • Hoang Cuong Nguyen successfully completed a CINTEL-funded Summer Vacation Project under the supervision of Mohan and Shahroz, focused on Cyber Threat Intelligence (CTI) visualisation. He is continuing this work through an Honours project.
  • Cedar Lett is an exceptional vacation scholarship student (PAWSEY student) (supervised by M.A.P. Chamikara, Mahathir Almashor (CSIRO Energy), and Marthie Grobler) working on devising an efficient approach to ensure privacy during DNA analysis. Recently, an update about Cedar’s experience at CSIRO’s data61 was shared at https://algorithm.data61.csiro.au/summer-of-science-meet-three-of-our-vacation-students/. Cedar is one of two Indigenous vacation students who have been supported by Data61 to participate in the Pawsey Supercomputing Research Centre’s Summer Internship Program over the past two years.
  • Akbar Fadiansyah has completed the 2022 Pawsey Supercomputing Research Centre’s Summer Internship Program under the supervision of Dr Chehara Pathmabandu and Dr Mahathir Almashor (CSIRO Energy). He was an outstanding and hardworking student who worked on both qualitative and quantitative research designs within a limited time exceeding our expectations. He conducted an extensive literature review on commercial IoT and privacy in developing a threat model, developed an anomaly detection system for time-series building IoT data and proposed a framework to mitigate privacy threats in commercial IoT systems. See https://megascrapper.github.io/pawsey-poster-site/
  • Hoang Cuong Nguyen successfully completed a CINTEL-funded Summer Vacation Project under the supervision of Mohan Baruwal Chhetri and Shahroz Tariq, focused on Cyber Threat Intelligence (CTI) visualisation. He will continue this work through an Honours project.
  • Aashima Jaiswal participated in the 2022 Pawsey Supercomputing Research Centre’s Summer Internship Program. Supervised by Tooba Aamir and Mahathir Almashor (CSIRO ENERGY), she worked on developing a hybrid ML/AI model to analyse social content for cyberbullying triggers and indicators. Her research focused on identifying the prevalence of cyberbullying as a significant issue for online communities and emphasised the need for automated tools for a safe cyberspace.
  • Welcome to Larry Huynh

‘I am a PhD student at the University of Western Australia, looking into developing quantum-inspired machine learning algorithms for improving cybersecurity. My research interests lie in this intersection of classical machine learning and quantum computing, and exploring the possible advantages that can be derived from quantum mechanics and distilled into classical computing models. I’ve also gained experience in various applications of classical machine learning. My work includes NLP for rumour generation to better combat the limitations of rumour detection models, anomaly detection techniques for maritime contexts with limited available data, and enhancing methods for visual field defect classification using perimetry data.’

Networking

Cyber Security CRC review