Skip to content

Imaging + AI for Graphics, Vision & Beyond

This research will focus on combining novel computational imaging methods with AI and deep learning techniques to address previously unsolvable or analytically intractable inverse problems in diverse fields such as realistic computer graphics, computer vision, signal processing, medical imaging and natural/physical sciences. Of particular interest is investigating novel techniques for multimodal imaging and AI based analysis of bio-physical properties of subsurface subject/material appearance for wide-ranging applications in computer graphics and vision and beyond. A key novel approach will be to combine practical, multimodal imaging of bio-physical appearance properties with differentiable inverse rendering and deep learning techniques to tackle hard inverse problems of appearance analysis from sparse, possibly incomplete/mixed observations.

Led by Dr Abhijeet Ghosh.

Our research