Prof. John Collomosse is a principal scientist in Adobe Research’s Imaging lab, and is based out of the London (UK) office.  John leads the Cross-modal Representation Learning (XRL) research group.

John has over 15 years experience at the intersection of Computer Vision, Graphics and AI, completing his PhD on Image and Video Stylization at the University of Bath in 2004.  John is concurrently a full professor at the University of Surrey, Centre for Vision Speech and Signal Processing (CVSSP) where he directs the DECaDE research centre.

John’s research focuses on representation learning for visual search and content discovery, focusing on creative modalities such as search, style and design layout.  His work has been demoed at Adobe MAX and appears in products e.g. Behance ‘More Like This’ style similarity search and Photoshop Content Credentials (image authenticity).  He has worked on volumetric video and performance capture for immersive AR/VR content production and video-realistic character animation.  John leads on research with Adobe’s Content Authenticity Initiative (CAI), researching content fingerprinting and decentralized technologies for visual content provenance.  He leads a task force on this topic for the C2PA cross-industry standards body on content authenticty.

John serves on the advisory boards of the UK Engineering and Science Council EPSRC and for the UKRI Digital Economy theme.