Per-clip and per-bitrate adaptation of the Lagrangian multiplier in video coding

TL;DR

This paper introduces a framework for adaptively selecting the Lagrangian multiplier in video coding across different bitrates, optimizing rate-distortion tradeoffs for individual clips to improve compression efficiency.

Contribution

It presents a novel method for per-operating point Lagrangian multiplier selection, enhancing rate-distortion optimization across bitrate ranges for individual videos.

Findings

Achieved approximately 2% BD-Rate improvements on HEVC videos.

Optimized Lagrangian parameters across multiple bitrates for 2,000 clips.

Improved rate-distortion performance using direct optimization techniques.

Abstract

In the past ten years there have been significant developments in optimization of transcoding parameters on a per-clip rather than per-genre basis. In our recent work we have presented per-clip optimization for the Lagrangian multiplier in Rate controlled compression, which yielded BD-Rate improvements of approximately 2\% across a corpus of videos using HEVC. However, in a video streaming application, the focus is on optimizing the rate/distortion tradeoff at a particular bitrate and not on average across a range of performance. We observed in previous work that a particular multiplier might give BD rate improvements over a certain range of bitrates, but not the entire range. Using different parameters across the range would improve gains overall. Therefore here we present a framework for choosing the best Lagrangian multiplier on a per-operating point basis across a range of bitrates. In effect, we are trying to find the para-optimal gain across bitrate and distortion for a single clip. In the experiments presented we employ direct optimization techniques to estimate this Lagrangian parameter path approximately 2,000 video clips. The clips are primarily from the YouTube-UGC dataset. We optimize both for bitrate savings as well as distortion metrics (PSNR, SSIM).