I am presently a fourth year postgraduate research student studying for a Doctor of Philosophy (PhD) degree in Computer Science at University of Warwick; I commenced the course in academic year 2013/14. Within the Department of Computer Science, I am working in the research domain of Methodologies & Applications. More specifically, I am a member of the Multimedia Processing & Computer Vision group working on High Efficiency Video Coding (HEVC) research.
Doctor of Philosophy (PhD) - Computer Science
The general research area on which I am presently focusing is the following video data compression standard: HEVC/H.265; this video coding platform has been standardised by JCT-VC (ITU-T/ISO/IEC: VCEG & MPEG). In my research undertakings, three original contributions to knowledge are proposed in the specialised field of adaptive quantisation for the HEVC standard. These original contributions fall into the following three categories:
1) HVS-JND adaptive perceptual CB level quantisation — one original contribution (C-BAQ & Adaptive Luminance-Chrominance JND Model).
2) HVS-CSF adaptive perceptual quantisation matrices — one original contribution (AQM).
3) Adaptive mathematical quantisation (i.e., scalar quantisation based purely on rate distortion theory) — one original contribution (AQST).
The HVS-JND and HVS-CSF contributions, named C-BAQ (including a novel luminance-chrominance JND model) and AQM, respectively, are designed for the purpose of exploiting the maximum levels of quantisation that can be applied to YCbCr input video data without incurring a palpable loss of perceptual reconstruction quality in the coded video, thereby potentially resulting in important improvements in terms of bitrate reduction while maintaining visual fidelity (as measured by BD-Rate, SSIM and ITU-T standardised subjective evaluations). Conversely, the adaptive mathematical quantisation contribution, named AQST, focuses on converting URQ from a non-adaptive, TB level technique into an adaptive, non-uniform transform coefficient level technique, whereby each coefficient is quantised individually based on its importance in the signal reconstruction process, thus leading to noteworthy improvements in visual fidelity while maintaining the same bitrate (as measured by BD-PSNR) — (Link: University of Warwick HEVC Research Profile).
The following papers are selected examples of the work I have undertaken as a PhD student.
PEER REVIEWED PAPERS
Lee Prangnell and Victor Sanchez, "Adaptive Quantization Matrices for HD and UHD Resolutions in Scalable HEVC," IEEE Data Compression Conference 2016, Utah, United States, 2016. (PDF) — AQM
Lee Prangnell, Victor Sanchez and Rahul Vanam, "Adaptive Quantization by Soft Thresholding in HEVC," IEEE Picture Coding Symposium 2015, Queensland, Australia, 2015. (PDF) — AQST
IEEE Copyright Notice
Publications are included on this website with permission from IEEE. Internal or personal use of this material is permitted. However, permission to reprint or republish this material for advertising or promotional purposes, or for creating new collective works for resale or redistribution, must be obtained from IEEE by writing to email@example.com. By choosing to view these documents, you agree to all provisions of the copyright laws protecting them. For more details, visit: http://www.ieee.org/publications_standards/publications/rights/copyrightpolicy.html.
WORKING & DISCUSSION PAPERS
Lee Prangnell and Victor Sanchez, "Color-Based Coding Unit Level Adaptive Quantization for HEVC," Working Paper, Mathematics Institute, University of Warwick, Coventry, UK, 2016. (PDF) — C-BAQ
Lee Prangnell, "Visible Light-Based Human Visual System Conceptual Model," Discussion Paper, Mathematics Institute, University of Warwick, Coventry, UK, 2016. (PDF) — C-BAQ
Master of Science (MSc) - Information Technology
The following is a tabulated summary of the final module grades with which I was awarded.
Final Grade: Master of Science (MSc) degree with Distinction (First Class).