Overview
Background
I am a passionate researcher with a background in computer science and a strong commitment to leveraging technology for the betterment of society. I hold a PhD in Image Forensics and have had the privilege of conducting postdoctoral research at prestigious institutions such as SUNY Albany and Dartmouth College, where I had the opportunity to collaborate closely with Prof. Hany Farid.
During my postdoc at Dartmouth College, I focused my research efforts on addressing a critical societal issue - real-time child pornography detection. This research not only garnered recognition within the academic community but also earned praise from luminaries like Prof. Ramesh Raskar at MIT, who invited me to share my insights through a talk at MIT.
I primarily works in the area of Cyber Security, Digital Forensics, Privacy and Security Aspects, Homomorphic Encryption and Cloud Computing.
As I continue my research journey, I remain committed to making a positive impact through innovation and collaboration. I am excited about the opportunities that lie ahead and the potential for technology to create a safer and more inclusive world.
Availability
- Dr Priyanka Singh is:
- Available for supervision
Fields of research
Qualifications
- Bachelor of Information Technology, Uttar Pradesh Technical University
- Masters (Research) of Image Processing, Motilal Nehru National Institute of Technology
- Doctor of Philosophy of Image Processing, Motilal Nehru National Institute of Technology
Research interests
-
Towards Mitigation of AI Based Attacks
The adoption of Artificial Intelligence (AI) has made a paradigmatic shift in the cybersecurity challenges. On the positive side, AI enables faster detection of anomalies, automated responses to attacks, and predictive threat modeling. This helps organisations to proactively combat cyber threats and minimise the impact of attacks. On the contrary, the integration of AI also opens new avenues for adversaries. Cybercriminals are increasingly leveraging AI to craft sophisticated attacks, such as automated phishing schemes, deepfake technology, and AI-driven malware that can adapt to traditional defence mechanisms. The existing cybersecurity frameworks lack adequate update to address this sophistication.
-
Facilitating confidential and effective Cyber Threat Intelligence sharing
Cyber Threat Intelligence (CTI) data is shared 24/7 between industry organisations and government agencies to provide incident responders real-time information about attackers' tactics and origins. Theoretically, this collaborative approach uplifts sectors’ cyber resilience but in reality, CTI sharing is often done inefficiently and ineffectively due to organisations' reticence to share full information which may unknowingly reveal organisational data (e.g., to competing or unknown entities within the community). This project aims to urgently address these challenges through a secure, collaborative CTI sharing approach with robust privacy protection built for AI-driven cybersecurity systems facilitating swift sector-wide incident response.
-
Security and Privacy of Large Language Models
As Large Language Models (LLMs) become increasingly pervasive in applications ranging from natural language understanding to creative content generation, the need for robust privacy protections becomes more critical. These models, which are trained on vast amounts of data, raise significant concerns regarding data security, user privacy, and ethical implications. This project aims to explore the privacy challenges inherent in LLMs, addressing both technical and legal dimensions while highlighting best practices for privacy-preserving AI development.
-
Combat Propagation of Dis/Misinformation on Social Media
Disinformation refers to deliberate misleading of people by manipulating narrative or facts. The unprecedented propagation of disinformation on social media can influence one’s mindset and cause an adverse impact on the society. AI-enabled tracking of the entire pipeline of disinformation by identifying the source, the target audience, the hidden agenda, the media of propagation, and the kind of influence can potentially combat this problem by reverse-identification of the origin of disinformation narratives and helping authorities to address the falsification problem.
-
From Prediction to Explanation: Multimodal, Explainable, and Interactive Deepfake Detection Framework for Non-Expert Users
The proliferation of deepfake technologies poses urgent challenges and serious risks to digital integrity, particularly within critical sectors such as forensics, journalism, and the legal system. While existing detection systems have made significant progress in classification accuracy, they typically function as black-box models-offering limited transparency and minimal support for human reasoning. This lack of interpretability hinders their usability in real-world decision-making contexts, especially for non-expert users.
Research impacts
My work focusses on using cutting-edge technology to tackle real-world challenges, particularly in the realm of online safety and security. I am driven by the belief that responsible and ethical technology can play a transformative role in protecting the most vulnerable in our society.
Works
Search Professor Priyanka Singh’s works on UQ eSpace
2016
Conference Publication
Region of Interest Based Robust Watermarking Scheme Exploiting the Homogeneity Analysis
Singh, Priyanka, Raman, Balasubramanian and Misra, Manoj (2016). Region of Interest Based Robust Watermarking Scheme Exploiting the Homogeneity Analysis. IEEE Region 10 Conference (TENCON), Singapore, 22-25 November 2016. Piscataway, NJ United States: Institute of Electrical and Electronics Engineers. doi: 10.1109/TENCON.2016.7848385
2016
Journal Article
An efficient fragile watermarking scheme with multilevel tamper detection and recovery based on dynamic domain selection
Singh, Priyanka and Agarwal, Suneeta (2016). An efficient fragile watermarking scheme with multilevel tamper detection and recovery based on dynamic domain selection. Multimedia Tools and Applications, 75 (14), 8165-8194. doi: 10.1007/s11042-015-2736-9
2016
Conference Publication
Automatic classification of leukocytes using morphological features and Naive Bayes classifier
Gautam, Anjali, Singh, Priyanka, Raman, Balasubramanian and Bhadauria, Harvendra (2016). Automatic classification of leukocytes using morphological features and Naive Bayes classifier. IEEE Region 10 Conference (TENCON), Singapore, 22-25 November 2016. New York, NY, United States: IEEE. doi: 10.1109/TENCON.2016.7848161
2014
Conference Publication
A chaotic map based DCT-SVD watermarking scheme for rightful ownership verification
Singh, Priyanka, Shivani, Shivendra and Agarwal, Suneeta (2014). A chaotic map based DCT-SVD watermarking scheme for rightful ownership verification. 2014 Students Conference on Engineering and Systems, Allahabad, India, 28-30 May 2014. Piscataway, NJ United States: Institute of Electrical and Electronics Engineers. doi: 10.1109/sces.2014.6880048
2013
Conference Publication
A hybrid DWT-SVD based robust watermarking scheme for color images and its comparative performance in YIQ and YUV color spaces
Singh, Priyanka, Agarwal, Suneeta and Pandey, Akanksha (2013). A hybrid DWT-SVD based robust watermarking scheme for color images and its comparative performance in YIQ and YUV color spaces. 3rd IEEE International Advance Computing Conference (IACC), Ghaziabad, India, 22-23 February 2013. New York, NY, United States: IEEE. doi: 10.1109/IAdCC.2013.6514400
2013
Conference Publication
A hybrid DCT-SVD based robust watermarking scheme for copyright protection
Singh, Priyanka and Agarwal, Suneeta (2013). A hybrid DCT-SVD based robust watermarking scheme for copyright protection. AFRICON Conference, Pointe aux Piments, Mauritius, 9-12 September 2013. New York, NY, United States: IEEE. doi: 10.1109/AFRCON.2013.6757709
2012
Conference Publication
A region specific robust watermarking scheme based on singular value decomposition
Singh, Priyanka and Agarwal, Suneeta (2012). A region specific robust watermarking scheme based on singular value decomposition. SIN '12: 5th International Conference of Security of Information and Networks, Jaipur, India, 25 - 27 October 2012. New York, NY United States: Association for Computing Machinery. doi: 10.1145/2388576.2388592
Supervision
Availability
- Dr Priyanka Singh is:
- Available for supervision
Before you email them, read our advice on how to contact a supervisor.
Available projects
-
Research projects
-
Investigating vulnerabilities of Split Learning/Federated Learning/Split-Fed Learning Models
-
Defenses to build Robust Deep Learning Models
-
Image/Audio/Video Forensic Techniques to Detect Doctored Images/Deep Fakes
-
Privacy Preserving Frameworks for Various Applications
-
Geo-hashing and Location Based Services
-
Supervision history
Current supervision
-
Doctor Philosophy
Privacy preservation with Large Language Models
Principal Advisor
Other advisors: Professor Xue Li
-
Doctor Philosophy
Understanding Privacy Concerns About Living in Smart Cities and Enhancing Privacy Awareness
Principal Advisor
Other advisors: Professor Xue Li
-
Master Philosophy
Effective and versatile multimodal fusion
Principal Advisor
Other advisors: Professor Xue Li
-
Doctor Philosophy
Advancing cyberworthiness via integration of AI and Model-Based Systems Engineering (MBSE)
Principal Advisor
Other advisors: Professor Ryan Ko
-
Doctor Philosophy
Weighted Ensembles for Different machine learning model that support non-data-sharing / vertical partition
Associate Advisor
Other advisors: Professor Xue Li
-
Doctor Philosophy
Identification of Disinformation Origins with Evidence-Based Justification
Associate Advisor
Other advisors: Professor Xue Li
-
Doctor Philosophy
Behaviour-Oriented Tracking of Disinformation
Associate Advisor
Other advisors: Professor Xue Li
-
Master Philosophy
Mapping Emerging Narratives Composed of False Information
Associate Advisor
Other advisors: Professor Xue Li
Media
Enquiries
For media enquiries about Dr Priyanka Singh's areas of expertise, story ideas and help finding experts, contact our Media team: