Graduate student John Murray of the Electrical and Computer Engineering (ECE) Department was one of the co-authors of the Best Student Paper Award from the Institute of Electrical and Electronics Engineers (IEEE) 2021 International Symposium on Multimedia, which took place in December.
Under the supervision of his advisor, ECE Professor Michael Zink, Murray worked on the research described in the winning paper, which in general dealt with improving 360-degree video streaming under challenging backhaul networking conditions.
Along with Zink and Murray from UMass Amherst, the other researchers in the winning paper were lead author Ayush Sarkar and his advisor Klara Nahrstedt from the University of Illinois Urbana-Champaign and Mallesham Dasari from Stony Brook University.
According to the five authors, 360-degree videos achieve unique visual immersions temporally and spatially, allowing users to view panoramic content by seamlessly altering the locations of their viewports through head-mounted displays and players amenable to 360-degree media. 360-degree video content is currently mushrooming with the proliferation of devices that facilitate virtual reality and augmented reality video traffic, which is expected to reach 4.02 exabytes per month this year.
However, as the authors point out, streaming panoramic videos also come with added challenges, which inspired the research described in their winning paper. For instance, these videos are of much higher resolution than 2D videos, causing greater bandwidth consumption when streamed.
According to the authors, this increased bandwidth utilization puts tremendous strain on the network capacity of the cloud providers streaming these videos. That’s where the trailblazing research in this winning paper comes in.
“In this paper,” say the five authors, “we introduce ‘Low Latency, Low Bandwidth, Optimized Super-Resolution Backhaul for 360-Degree Video Streaming’ (or L3BOU), a novel, three-tier, distributed software framework that reduces cloud-edge bandwidth in the backhaul network and lowers average end-to-end latency for 360-degree video streaming applications.”
The authors add that “The L3BOU framework achieves low bandwidth and low latency by leveraging edge-based, optimized, upscaling techniques.”
The authors go on to explain that L3BOU accomplishes all these pioneering improvements by using a method known as down-scaled MPEG-DASH encoding, which is an adaptive bitrate streaming technique that enables streaming of media content over the Internet delivered from conventional HTTP web servers, and applying such encoding to 360-degree video data.
Then the L3BOU edge takes the resulting ultra-low-resolution data and utilizes distributed, super-resolution techniques, thus providing high-quality video to the client.
As the authors summarize their research, “L3BOU is able to reduce the cloud-edge backhaul bandwidth by up to a factor of 24, and the optimized super-resolution multi-processing of ULR data provides a 10-fold latency decrease in super-resolution upscaling at the edge.”