Bashayer Abdallah

I am a PhD candidate under the supervision of Dr Shan E Ahmed Raza, specialising in monocular depth estimation. My research focuses on developing advanced methods for predicting depth from single RGB images, with a particular emphasis on enhancing edge preservation and structural consistency within depth maps. By integrating edge-aware techniques, my work aims to improve the accuracy, robustness, and visual quality of monocular depth estimation models.

My PhD is funded by the Kuwait Institute for Scientific Research - KISRLink opens in a new window.

Email: bashayer.abdallah@warwick.ac.uk

GitHub | LinkedIn Profile

VISION Lab, TIA Centre, Computer Science
University of Warwick

Monocular Depth Estimation & recovery of accurate 3D structure from a single image, tackling absolute scale across different depth ranges and advancing prediction through diffusion-based refinement of sparse outputs, with applications in robotics, endoscopy, and autonomous systems.

Vision–Language Integration & Multimodal Learning. My work on DepthClassNet (MIUA 2025) introduced a multitask vision–language framework that couples monocular depth estimation with language-conditioned texture classification via CLIP-aligned cross-modal representations — demonstrating how NLP-guided inference can enrich perceptual models beyond purely visual supervision.

Biomedical Imaging & Clinical AI. Working on colonoscopic and endoscopic image analysis, where I have developed lightweight, deployment-ready models suited to real-time, resource-constrained clinical environments. I am broadly interested in AI for minimally invasive surgery, endoscopy, and medical image understanding.

Interdisciplinary & Scientific ML. Applying data-driven machine learning to domains beyond standard vision benchmarks.

• Associate Fellow of the Higher Education Academy (AFHEA) — Advance HE
  Awarded via Warwick APP PGR, aligned with the UK Professional Standards Framework (UKPSF).

• Graduate Student Member of the Institute of Electrical and Electronics Engineers (IEEE).

• Best Presentation Award, Machine Learning & AI Track, Warwick Postgraduate Colloquium in Computer Science (2023) - WPCCSLink opens in a new window

• Abdallah, B., Raza, S. E. A. Monocular Depth Estimation with Guided Edge-Aware Attention for Endoscopic Images. IEEE International Symposium on Biomedical Imaging (ISBI), 2026.

• Abdallah, B., Raza, S. E. A., Sanchez, V. Edge-guided Monocular Absolute Depth Estimation with Diffusion-based Refinement. IEEE International Conference on Image Processing (ICIP), 2025.

👉View Paper

• Abdallah, B., Raza, S. E. A. DepthClassNet: A Multitask Framework for Monocular Depth Estimation and Texture Classification in Endoscopic Imaging. Medical Image Understanding and Analysis (MIUA), Springer Lecture Notes in Computer Science, pp. 230–246, 2025.

👉View Paper Code: Bashayerq8/DepthClassNet

• Research Associate in the Petroleum Research Centre PRC at KISR.Link opens in a new window
• PRC Computer Lab supervisor since 2016.
• Participated in software development and ML projects for upstream and downstream oil enhancement.
• Senior Graduate Teaching Assistant at the University of Warwick:

  • CS342 Machine Learning - 2022 and 2023Link opens in a new window

  • CS133 Professional Skills - 2022, 2023 and 2024Link opens in a new window
  • CS355 Digital Forensics - 2023 and 2024Link opens in a new window
  • CS346 Advanced Databases - 2025
  • CS352 Project Management for Computer Scientists - 2025

- PhD in Computer Science (Expected summer 2026)

University of Warwick, UK

Dissertation: Edge-Aware and Multimodal Frameworks for Monocular Depth Estimation Across Absolute Depth Ranges

- MSc Advanced Software Technology — Distinction (2020)

Swansea University, United Kingdom

Dissertation: Geometric Convolutional Neural Networks for 2D Image Segmentation

- BSc Information Technology and Computing — Upper Second Class (2:1) (2015)

Arab Open University, Kuwait