Funding:
Home (UK) and EU citizens who have confirmation of UK settlement or pre-settlement status under the EU Settlement Scheme
The project:
Acquiring high-quality footage in challenging environments such as low light, heat haze, and adverse weather is significantly difficult. These conditions not only degrade video quality but also complicate interpretation by humans and machines, making post-processing crucial. However, video restoration and enhancement are complex due to information loss and the lack of ground truth data.
This project addresses these issues innovatively. We propose using prior information from high-quality videos that share content with distorted footage as constraints in the learning process of modelling algorithms. This method uses the characteristics and knowledge embedded in high-quality videos to guide the restoration and enhancement of distorted videos.
Our goal is to develop a comprehensive framework for video restoration by tackling blind inverse problems with unsupervised learning. The project involves a collaborative team, including a postdoctoral researcher and a PhD student, with specific objectives:
- Define and acquire a comprehensive database of high-quality video priors for reference in enhancing distorted videos.
- Develop a robust high-level representation of video content. This involves minimizing the gap between characteristics of input and high-quality reference videos in the database to maximize the accuracy of acquired priors.
- Create a prior retrieval system that provides global, local, and context-based priors, along with statistically driven models for the video restoration process.
- Address blind inverse problems by defining a network to learn distortion functions from data, informing the optimization in the learning process.
- Refine optimization and learning strategies that are aware of the acquisition context and capable of learning without explicit ground truth, using unsupervised learning to enhance video quality.
- These initiatives aim to significantly improve video quality, overcoming challenges presented by environmental conditions in footage acquisition.
How to apply:
Prior to submitting an online application, you will need to contact the project supervisor to discuss.
Online applications are made at http://www.bris.ac.uk/pg-howtoapply. Please select PhD in Computer Science on the Programme Choice page. You will be prompted to enter details of the studentship in the Funding and Research Details sections of the form.
Candidate requirements:
Applicants must hold/achieve a minimum of a merit at master’s degree level (or international equivalent) in a science, mathematics or engineering discipline. Applicants without a master's qualification may be considered on an exceptional basis, provided they hold a first-class undergraduate degree. Please note, acceptance will also depend on evidence of readiness to pursue a research degree.
If English is not your first language, you need to meet this profile level: Profile E
Further information about English language requirements and profile levels.
Funding:
Full fees and a tax-free stipend will match the UKRI rate. As a guide, for 2024/25 this is £19,237.
Contacts:
For questions about the research topic, please contact:
Dr Pui Anantrasirichai, N.Anantrasirichai@bristol.ac.uk
For questions about eligibility and the application process please contact Engineering Postgraduate Research Admissions admissions-engpgr@bristol.ac.uk
Full fees and a tax-free annual stipend will match the UKRI rate. As a guide, for 2024/25 this is £19,237.
