Real-World Video Adaptation with Reinforcement Learning
Abstract
Client-side video players employ adaptive bitrate (ABR) algorithms to optimize user quality of experience (QoE).We evaluate recently proposed RL-based ABR methods in Facebook's web-based video streaming platform. Real-world ABR contains several challenges that requires customized designs beyond off-the-shelf RL algorithms\,---we implement a scalable neural network architecture that supports videos with arbitrary bitrate encodings; we design a training method to cope with the variance resulting from the stochasticity in network conditions; and we leverage constrained Bayesian optimization for reward shaping in order to optimize the conflicting QoE objectives. In a week-long worldwide deployment with more than 30 million video streaming sessions, our RL approach outperforms the prior ABR algorithm.
Cite
Text
Mao et al. "Real-World Video Adaptation with Reinforcement Learning." ICML 2019 Workshops: RL4RealLife, 2019.Markdown
[Mao et al. "Real-World Video Adaptation with Reinforcement Learning." ICML 2019 Workshops: RL4RealLife, 2019.](https://mlanthology.org/icmlw/2019/mao2019icmlw-realworld/)BibTeX
@inproceedings{mao2019icmlw-realworld,
title = {{Real-World Video Adaptation with Reinforcement Learning}},
author = {Mao, Hongzi and Chen, Shannon and Dimmery, Drew and Singh, Shaun and Blaisdell, Drew and Tian, Yuandong and Alizadeh, Mohammad and Bakshy, Eytan},
booktitle = {ICML 2019 Workshops: RL4RealLife},
year = {2019},
url = {https://mlanthology.org/icmlw/2019/mao2019icmlw-realworld/}
}