ESR8 – Tom Gillooly

Tom Gillooly

Country of Origin: UK/New Zealand
Host Institution: NTNU

Background:

Tom has a bachelor's degree in Electrical and Electronic Engineering from the University of Auckland, New Zealand. He later pursued a Master in Applied Computer Science at NTNU i Gjovik, under the COSI program. He has five years commercial experience as an embedded software engineer, and worked as an intern at Disney Research Zurich in the Projection Technologies team. Prior to joining the ApPEARS project, he was a research engineer at INSA Lyon, researching Vision-Language Navigation.

Aspirations within projects:

Tom is interested in finding the cues that our visual system uses to recognize different materials without us realizing it. If these cues are identified, can we exploit them to accurately simulate material appearance? How would we even know if we've done this successfully? Answering these questions demands a real multi-disciplinary approach, spanning computer vision, imaging, computer graphics, human perception, and art. He hopes to synthesize these domains and discover what each can contribute to the others. He'll be coming into the project of the back of a year and half of deep learning research, and is curious as to how modern machine learning techniques can be integrated with the material appearance rendering pipelines of the future.

ESR8: 3D soft proofing and appearance simulation

Main Supervisor:

Co-Supervisor(s):

Objectives

Combine knowledge from computer graphics, vision, and imaging to define new methods for providing accurate visual previews of 2.5D and 3D printed material appearance.

  • Identify major physical and perceptual attributes that correlate to the visual appearance of printed surfaces.
  • Address specifically surface roughness and the appearance of gloss, in conjunction with colour.
  • Develop enabling technologies and methodologies for visualization of material appearance.
  • Evaluate quality of developed technologies and methodologies using human observers and appropriate statistical methods.