Content-gnostic Bitrate Ladder Prediction for Adaptive Video Streaming

Research output: Chapter in Book/Report/Conference proceedingConference Contribution (Conference Proceeding)

21 Citations (Scopus)
380 Downloads (Pure)


A challenge that many video providers face is the heterogeneity of networks and display devices for streaming, as well as dealing with a wide variety of content with different encoding performance. In the past, a fixed bit rate ladder solutionbased on a ”fitting all” approach has been employed. However, such a content-tailored solution is highly demanding; the computational and financial cost of constructing the convex hull per video by encoding at all resolutions and quantization levels is huge. In this paper, we propose a content-gnostic approachthat exploits machine learning to predict the bit rate rangesfor different resolutions. This has the advantage of significantlyreducing the number of encodes required. The first results, based on over 100 HEVC-encoded sequences demonstrate the potential, showing an average Bjøntegaard Delta Rate (BDRate) loss of 0.51% and an average BDPSNR loss of 0.01 dB compared to the ground truth, while significantly reducing the number of pre-encodes required when compared to two other methods (by 81%-94%).
Original languageEnglish
Title of host publicationPicture Coding Symposium 2019
PublisherIEEE Computer Society
Number of pages5
ISBN (Electronic)978-1-7281-4704-8
ISBN (Print)978-1-7281-4705-5
Publication statusPublished - 9 Jan 2020
EventPicture Coding Symposium 2019 - Ningbo, China
Duration: 12 Nov 201915 Nov 2019
Conference number: 34

Publication series

NamePicture Coding Symposium (PCS)
ISSN (Print)2330-7935
ISSN (Electronic)2472-7822


ConferencePicture Coding Symposium 2019
Abbreviated titlePCS2019
Internet address


  • Rate-Quality Convex Hull
  • Bitrate Ladder
  • Pertitle Video Encoding
  • HEVC
  • Adaptive Video Streaming


Dive into the research topics of 'Content-gnostic Bitrate Ladder Prediction for Adaptive Video Streaming'. Together they form a unique fingerprint.

Cite this