Visible but imperceivable
Generative AI can produce beautiful and useful image content for a variety of applications. The identification of AI-produced content is important for both humans and future generative AI models. This project seeks to create a hyper-robust way of actively watermarking generated content by creating visible yet commonly imperceivable watermarks.
Research Opportunities
Summary
As AI-generated content becomes ever more common in our digital world, the challenge of differentiating between human- and AI- produced images rises both in importance and difficulty. Previously, this has been done using image watermarking and steganography which have been used as a way of embedding within the pixels details of an image’s history or context. Traditionally, steganography techniques have balanced robustness and visibility. Actively changing the visible content of an image has been seen to be undesirable for a variety of reasons. However, the less visible a watermark is, the more prone it tends to be against removal attacks such as compression or transformation, or even just accidental removal. Compression methods tend to target biological, bottom-up weaknesses in the human visual system, thus removing redundances in the images that human eyes cannot see, and subsequently removing any watermarks or steganography content hidden within these features. At the same time, optical illusions demonstrate that humans commonly have some cognitive and perceptual weaknesses that render them blind to some image effects. Examples include gradual change blindness (“The Changing Room Illusion” by Michael Cohen) and differences in image saliency and selective attention, including that related to top-down processing and ambigrams.
Aim
To create a methodology and variety of techniques that will produce robust watermarks for images/videos by making watermarks that are easily visible/detectable with the naked eye if there is prior knowledge of their existence, but otherwise imperceptible to the casual viewer.
Objectives
- To explore and critically analyse what facets of human vision and cognition contribute to an image or image transform being perceivable or imperceivable.
- To produce and a visible-yet-imperceivable watermarking methodology for digital images and a subsequent reliable computer vision/machine learning classification/detection method for the presence of such watermarks. The limits of the methodology should be examined as it may be unlikely that a completely imperceivable yet visible watermark exists.
- To evaluate the robustness of the novel watermarking methods against common removal attacks and/or image processing techniques that disrupt watermarks (e.g. rebroadcast attacks, affine transforms, targeted compression)
- To investigate the relationship between image content and context to imperceivable watermarks i.e. how does the watermark have to change to adapt to different types of images and different contexts such as photographs, artwork, medical images etc.
Candidates
Candidates should have some knowledge of digital image creation/processing, either using existing image editing/generation tools (e.g. Adobe suite, StableDiffusion etc.) or programmatically (e.g. Python, OpenCV). Familiarity with computer vision techniques as well as a background in machine learning would be considered an advantage (e.g. TensorFlow, Keras). The project is also likely to involve user studies in order to quantify perception of proposed watermarking methods.
Supervisors
Discuss this further with a potential supervisor for this research degree:
Research Themes
Find other Research Degrees in the same theme:
About our Research Degrees
We welcome specific ideas from applicants for a potential research degree topic. The University offers a number of potential self-funded topics in each School which you could consider developing your research project around.
It is not unusual for a proposed research degree topic to evolve as you progress through the application stage and even your future studies. This ensures the best match of your academic/professional experience with that of your supervisory team.
Occasionally, the University will also advertise available funded studentships, for very specific topics or subjects:
If you need any further support prior to applying, please email researchdegrees@rgu.ac.uk
Entry requirements
Entry Requirements for Research Degrees
Normal entry requirements are a first or upper second class honours degree from a University of the United Kingdom, or from an overseas University, or degree equivalent qualification, subject to the approval of Robert Gordon University.
Applicants holding other qualifications shall be considered on their merits and in relation to the nature and scope of the proposed research programme.
Applicants will normally register for an MSc/PhD or MRes/PhD with transfer to PhD dependent on satisfactory progress. However, applicants who already possess a Master’s degree which includes a significant research component could be registered directly for a PhD, subject to a satisfactory research proposal, and clear evidence of research expertise.
Some Research Degree Topics require employment experience, research experience and/or transcripts for international qualifications.
Contact researchdegrees@rgu.ac.uk to see if you are eligible.
English Language
Any applicant whose first language or language of previous University-level instruction is not English must normally demonstrate competence in English. This should be a score of IELTS 6.5 with 6.5 in all elements, or an approved equivalent.
We accept a variety of in-country and secure English language tests, find out more:
Please note, some courses may require a higher standard of English than stated in this page. Contact researchdegrees@rgu.ac.uk for further information.
Fees & Costs
For Academic Year 2024/2025
UK/EU Students
- Full time - £4,786 per academic year
- Part time MPhil/PhD - £2,870 per academic year
- Part time MRes/MSc by research - £2,390 per academic year
International Students
- Full time - £19,450 per academic year
- Part-time MPhil/PhD - £11,670 per academic year
- Part-time MRes/MSc by research - £9,720 per academic year
For Academic Year 2023/2024
UK/EU Students
- Full time - £4,712 per academic year
- Part time MPhil/PhD - £2,830 per academic year
- Part time MRes/MSc by research - £2,356 per academic year
International Students
- Full time - £18,700 per academic year
- Part-time MPhil/PhD - £11,220 per academic year
- Part-time MRes/MSc by research - £9,350 per academic year
Additional Costs
The following course-related costs are not included in the course fees:
- The cost of books that you may wish to purchase
- Accommodation and living costs
- Printing
Alumni Discount
Robert Gordon University is delighted to offer a 20% discount on course fees for all alumni who have graduated from RGU.
Scholarships and Financial Awards
Postgraduate research students will normally make their own arrangements for payment of fees. However you may qualify for a scholarship or financial award:
Funded Studentships
Occasionally, the University will also advertise available funded studentships, for very specific topics or subjects:
Disclaimer
For new intakes course fees are reviewed and published annually for each mode of delivery. Tuition fees are fixed for the duration of a course at the rate confirmed in the offer letter. For further information see:
How to Apply
Applications
All applications should be made via the University's online application for research students.
Before you apply, find out more about the application process, research proposal writing guides, routes and durations:
Our Research degrees usually start in February and October each year. There are no formal deadlines for you to submit your application, though we recommend you submit it as early as possible in order to ensure we can help you join us at your preferred time.
Completed application forms should be accompanied by:
- Degree certificate
- Statement of purpose, detailing why you would like to study this degree in the UK, and why you have chosen RGU.
- A short research proposal, in the range of 1,500 - 3,000 words, containing a rationale and background, study aims and objectives, a literature review, proposed research methods and a full reference list.
Further guidance on the application process:
English Language Requirements
All applicants whose first language is not from a recognised English speaking country will be asked to provide evidence of their English language skills as part of their offer to study at RGU.
Find out more about the English language qualification we accept:
Ready to start this Research Degree?
Find out about our entry requirements, application dates and how to apply.
Any questions?
Get in touch with our team and we'll do our best to help.