Sign In

Communications of the ACM

ACM TechNews

UCI, Disney Research Scientists Develop AI-Enhanced Video Compression Model

View as: Print Mobile App Share:
University of California, Irvine assistant professor Stephan Mandt

University of California, Irvine assistant professor Stephan Mandt says every video compression approach "works on a trade-off."

Credit: Steve Zylius / UCI

An artificial intelligence (AI)-enhanced video compression model developed by computer scientists at the University of California, Irvine (UCI) and Disney Research has shown that deep learning can provide results comparable to established video compression technology.

The team showed its compressor yielded less distortion and significantly smaller bits-per-pixel rates than classical coding-decoding algorithms when trained on specialized video content.

The compressor achieved similar results on downscaled, publicly available YouTube videos. “Ultimately, every video compression approach works on a trade-off,” said UCI's Stephan Mandt. "The hope is that our neural network-based approach does a better trade-off overall between file size and quality.”

From University of California
View Full Article


Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account