A research team from the Bristol Vision Institute (BVI) has been awarded a key technology prize for its work on video compression.

“The IEEE prize is a clear recognition of the strength of Bristol’s compression research”

 

The ViSTRA technology compresses video in a way that is less noticeable to the human eye across each frame and also between frames. This ‘perceptually optimised spatio-temporal resampling’ is included in the video encoder and is combined with a neural network-based method of reconstructing the output video that understands the content in each frame.

This combination dramatically reduces bit rate and increases visual quality. The codec has attracted considerable attention from industry and the team is currently preparing a contribution to the next generation of MPEG video coding standards beyond the current HEVC standard.

Professor David Bull, Dr Fan Zhang and PhD student Mariana Afonso were awarded the prize at the ICIP Conference held in Beijing, China in September, with PhD student Mariana Afonso receiving the prize for Video Coding based on Spatial Resolution Adaptation from Anne Aaron and Ioannis Katsavounides from Netflix.

“ViSTRA is the culmination of many years work in perceptually optimised video coding and dynamic resampling and the IEEE prize is a clear recognition of the strength of Bristol’s compression research that spans engineering, computer science and psychology,” said Professor David Bull, Director of the Bristol Vision Institute.

The Bristol Vision Institute is hosting the 2017 Richard Gregory Memorial Lecture on the The Science of Seeing Differently on Friday 3 November at 6pm at the School of Experimental Psychology, Lecture Theatre 1, Priory Road Complex

Nick Flaherty