Neural Adaptive Video Streaming with Pensieve


Hongzi Mao      Ravi Netravali      Mohammad Alizadeh     

Computer Science and Artificial Intelligence Laboratory
Massachusetts Institute of Technology


Abstract


Pensieve is a system that generates ABR algorithms using reinforcement learning. Pensieve trains a neural network model that selects bitrates for future video chunks based on observations collected by client video players. Pensieve does not rely on pre-programmed models or assumptions about the environment. Instead, it learns to make ABR decisions solely through observations of the resulting performance of past decisions. As a result, Pensieve automatically learns ABR algorithms that adapt to a wide range of environments and QoE metrics. We compare Pensieve to state-of-the-art ABR algorithms using trace-driven and real world experiments spanning a wide variety of network conditions, QoE metrics, and video properties. In all considered scenarios, Pensieve outperforms the best state-of-the-art scheme, with improvements in average QoE of 12%–25%. Pensieve also generalizes well, outperforming existing schemes even on networks for which it was not explicitly trained.


Video




Paper


Neural Adaptive Video Streaming with Pensieve
Hongzi Mao, Ravi Netravali, Mohammad Alizadeh
Proceedings of the 2017 ACM SIGCOMM Conference
[PDF]


Code


[GitHub]


Slides


[Slides]


Press


Forbes, Tech Crunch, Engadget, Daily Mail, Science Daily, MIT News, International Business Times, Electronic 360, Inverse, Digital Journal, FossBytes, and other media outlets covered Pensieve.



Supporters


This project is supported by NSF, the MIT Center for Wireless Networks and Mobile Computing, and a Qualcomm Innovation Fellowship.


More


Extended technical report and more of the project are coming soon