Home ▻ Research Centres▻College of Science and Engineering▻C3I▻Researchers▻Graduated Ph.D. Students▻Faisal Khan

Faisal Khan

Biography & Research Interests

I was born and raised in the Barawal Bandi Upper Dir district of Pakistan's Khyber Pakhtunkhwa province. From 2011 to 2015, I received a bachelor in mathematics from the University of Malakand, and from 2015 to 2017, I obtained a master's degree in mathematics from Hazara University Mansehra. I am currently a PhD student at Professor Peter Corcoran group since 2019 in College of Science and Engineering, National University of Ireland Galway (NUIG). My research areas of expertise include Depth Estimation, Facial Depth Estimation, 3D Reconstruction, Deep Learning, Machine Learning, Mathematical Modelling, Chemical Kinetics, Fuzzy sets, Decision Making, Optimization and Tensors analysis. My time at C3 Imaging Center has not only provided me with the knowledge and skills I need to manage my PhD more effectively and efficiently, but it has also given me a sense of personal satisfaction and the confidence to be anything I want to be.

Research Activities

•The process of estimating depth from a scene's image data is a crucial task in computer vision. It's especially important in image perception, where determining the primary objects and regions within an imaged scene, as well as their relative positions and orientations from frame to frame, can provide useful information about scene behaviour. Human faces are one of the most popular features in photographs, and they are a key part of major of image comprehension issues. Since the eye-separation in a human face is well-known in human anthropometry, it is possible to calculate the distance-to-camera of a human subject from a single image frame with considerable accuracy given knowledge of the camera's field-of-view. This research implies that by training a neural computer vision models on data containing human faces, it should be possible to learn a more accurate depth estimation. With enough data and pixel-accurate ground truth (GT), the models should be able to learn many nuances of human facial features and structure, improving depth estimation over the current SoA.  The key research activities includes:
 
1.Monocular Depth Estimation Methods based on Deep Learning
 
2.Facial Depth Models Generation (Synthetic data)
  • Blinder, iClon 7, iClone Character Creator
  • Using CNN and transfer learning methods to validate the generated facial depth models
  • A new Encoder-decoder model and their comparison with SoA
  • A study on Depth datasets and loss functions
  • Monocular Facial depth Neural networks, past, present and future

Journal Publications

Faisal Khan, Shahid Hussain, Shubhajit Basak, Joseph Lemley, Peter Corcoran. An efficient encoder–decoder model for portrait depth estimation from single images trained on pixel-accurate synthetic data, Neural Networks, Volume 142, 2021, Pages 479-491, ISSN 0893-6080, https://doi.org/10.1016/j.neunet.2021.07.007.

Faisal Khan, Muhammad Ali Farooq, Waseem Shariff, Shubhajit Basak, And Peter Corcoran, Towards Monocular Neural Facial Depth Estimation: Past, Present, and Future, in IEEE Access.

Faisal Khan, Shahid Hussain, Shubhajit Basak, Mohamed Moustafa, and Peter Corcoran (2021). "A Review of Benchmark Datasets and Training Loss Functions in Neural Depth Estimation," in IEEE Access, doi: 10.1109/ACCESS.2021.3124978.

Faisal Khan; Salahuddin, Saqib; Javidnia, Hossein. 2020. "Deep Learning-Based Monocular Depth Estimation Methods—A State-of-the-Art Review" Sensors 20, no. 8: 2272.

S. Basak, P. Corcoran, Faisal Khan, R. Mcdonnell and M. Schukat, "Learning 3D Head Pose from Synthetic Data: A Semi-Supervised Approach," in IEEE Access, vol. 9, pp. 37557-37573, 2021.

Conference Participation & Presentations

Faisal Khan, Shubhajit Basak and Peter Corcoran, "Accurate 2D Facial Depth Models Derived from a 3D Synthetic Dataset," 2021 IEEE International Conference on Consumer Electronics (ICCE), 2021, pp. 1-6.

Faisal Khan, S. Basak, H. Javidnia, M. Schukat and P. Corcoran, "High-Accuracy Facial Depth Models derived from 3D Synthetic Data," 2020 31st Irish Signals and Systems Conference (ISSC), 2020, pp. 1-5.

S. Basak, Faisal Khan, R. McDonnell and M. Schukat, "Learning Accurate Head Pose for Consumer Technology From 3D Synthetic Data," 2021 IEEE International Conference on Consumer Electronics (ICCE), 2021, pp. 1-6.

S. Basak, H. Javidnia, Faisal Khan, R. McDonnell and M. Schukat, "Methodology for Building Synthetic Datasets with Virtual Humans," 2020 31st Irish Signals and Systems Conference (ISSC), 2020, pp. 1-6.