Live Video Streaming Optimization Based on Deep Reinforcement Learning

Zhang, Xueshuai; Hu, Yuxiang; Li, Ziyong

doi:10.1145/3383972.3384058

Cited by 4 publications

(6 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is an indirect way to select the optimal strategy. Taking the update method of the Q-learning value function as an example [23]:…”

Section: B Qoe-oriented Deep Reinforcement Learningmentioning

confidence: 99%

A Deep Reinforcement Learning-Based Optimal Transmission Control Method for Streaming Videos

Yang,

Xiao

2024

IEEE Access

View full text Add to dashboard Cite

Time delay and image quality degradation are main challenges faced by streaming video transmission. How to make adaptive planning for transmission schemes according to dynamic change of transmission environment, always remains a technical concern. As a consequence, this paper proposes a deep reinforcement learning-based optimal transmission control method for streaming videos. Firstly, edge buffer task allocation is combined with quality of experience (QoE)-oriented deep reinforcement learning algorithm, in order to develop a resource allocation method for streaming videos. Secondly, an actively coordinated streaming data streaming transmission mechanism is established to construct a specific optimal transmission control method that satisfies environment requirement. Finally, a set of experiments are conducted to verify effectiveness and performance on public video transmission datasets. And the proposal is compared with several traditional transmission methods. The experiments show that the proposal in this work can effectively reduce delay and startup time and improve the QoE. This shows that the proposal is able to bring better stability and transmission quality.

show abstract

“…This is an indirect way to select the optimal strategy. Taking the update method of the Q-learning value function as an example [23]:…”

Section: B Qoe-oriented Deep Reinforcement Learningmentioning

confidence: 99%

A Deep Reinforcement Learning-Based Optimal Transmission Control Method for Streaming Videos

Yang,

Xiao

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Video viewing and live punch-in are automatically counted and calculated by the system. The training teacher's score, the same group of students' score and the post unit's score are based on Wilson's theorem, as shown in Formula 1 [8].…”

Section: Training Evaluationmentioning

confidence: 99%

Design and Development of Remote Training Platform for Work-Study Skills Based on Streaming Media

Ding¹

2023

Atlantis Highlights in Social Sciences, Education and Humanities

View full text Add to dashboard Cite

In order to solve the problems of difficulty, high cost, fixed time and limited space in the remote training of work-study skills in China, this paper takes the work-study training as the research object, and designs a set of remote training platform for work-study skills by combining the technologies of network information, database management and computer application. The platform provides a necessary help for optimizing the work-study process, improving students' comprehensive quality and ensuring the smooth development of the work-study program in schools.

show abstract

“…In 2019 Suphx, made by MSRA, beat most of the top human players in Riichi Mahjong [18], and in 2022 JueJong, built by Tencent AI Lab, beat a human champion in two-player Mahjong [19]. There are also efforts to solve Doudizhu, a popular card game in China, including DouZero [20] and PerfectDou [21], but no superhuman performance has yet been reported.…”

Section: Game Ai Benchmarksmentioning

confidence: 99%

“…Each of these game AI systems uses different combinations of AI techniques listed in Section 3. The AI systems covered in this section are: AlphaGo series (AlphaGo [2], Al-phaGo Zero [3], AlphaZero [4]) in the game of Go, HUNL AI systems (DeepStack [5], Libratus [6], Pluribus [7], AlphaHoldem [80]), Mahjong AI systems (Suphx [18], JueJong [19]), Doudizhu AI systems (DouZero [20], PerfectDou [21]), and AI systems for video games (AlphaStar [8], OpenAI Five [9], JueWu [10]).…”

Section: Milestones Of Game Ai Systemsmentioning

confidence: 99%

Techniques and Paradigms in Modern Game AI Systems

2022

Algorithms

View full text Add to dashboard Cite

Games have long been benchmarks and test-beds for AI algorithms. With the development of AI techniques and the boost of computational power, modern game AI systems have achieved superhuman performance in many games played by humans. These games have various features and present different challenges to AI research, so the algorithms used in each of these AI systems vary. This survey aims to give a systematic review of the techniques and paradigms used in modern game AI systems. By decomposing each of the recent milestones into basic components and comparing them based on the features of games, we summarize the common paradigms to build game AI systems and their scope and limitations. We claim that deep reinforcement learning is the most general methodology to become a mainstream method for games with higher complexity. We hope this survey can both provide a review of game AI algorithms and bring inspiration to the game AI community for future directions.

show abstract

Live Video Streaming Optimization Based on Deep Reinforcement Learning

Cited by 4 publications

References 4 publications

A Deep Reinforcement Learning-Based Optimal Transmission Control Method for Streaming Videos

A Deep Reinforcement Learning-Based Optimal Transmission Control Method for Streaming Videos

Design and Development of Remote Training Platform for Work-Study Skills Based on Streaming Media

Techniques and Paradigms in Modern Game AI Systems

Contact Info

Product

Resources

About