Following the Beijing Multimedia Development Salon on March 31, LiveVideoStack entered Shanghai, hoping to show the exploration and practice of new technology in the field of audio and video, as well as the latest and best practices of emerging application scenarios and traditional industries through the “New Trend of multimedia Development” series salon.


With the blooming of Codec, the finalization of WebRTC1.0, the rapid development of AI and blockchain technology, we have seen the integration and practice of multimedia and more new technologies. On April 21, we invited six technology masters from Tencent, Youku, Touch Treasure Technology, SUNing PP Video Cloud, Trisaran cloud and Xiangxin Technology to share the technology practice and future outlook of multimedia development.


Lecturers and Topics


Wu Weiqi touch Treasure technology audio technology expert

Speaker: Audio technology expert of Touch Treasure, responsible for echo cancellation, voice denoising, voice codec, voice quality optimization and other aspects of touch Treasure mobile phone. Before that, he worked in iQiyi, Ali Music and student Jun with excellent performance as audio algorithm. Graduated from The School of Communication and Information Engineering, Shanghai University in 2013, majoring in Signal and Information Processing, master degree. Published two papers and a number of patents.

Topic: Audio post-processing technology of real-time voice communication

In the process of real-time voice communication, noise, reverberation and echo are inevitably introduced, which greatly reduces the voice quality and seriously interferes with the normal conversation of users. Therefore, audio post-processing technology is very important. In view of these problems, this speech will mainly focus on echo cancellation, noise suppression, automatic volume control three parts of content sharing, and the future development of audio post-processing technology.


Zeng Xiaowei Deputy Technical Director of SUNing IT Headquarters

Speaker: Now I am deputy director of PP video cloud technology of Suning Business Data Cloud Company. Image codec, high performance computing background; Engaged in AI (NLP) direction, more than 10 years of streaming media service server-side development and architecture design experience, more than 8 years of team management experience. Currently, I am responsible for the architecture design and research and development of PP video cloud, undertake the internal video business of Suning Group, promote the improvement of video cloud platform architecture and enhance service capabilities; Externally, it provides high-quality video SAAS services. At the same time, he is a writer observer of many science and technology websites, and a long-term guest writer and executive editor of PP video cloud.

Topic: How to use machine learning algorithms to optimize distribution links

After the voD video (including short video) is pressed, it can be connected to the VIDEO CDN only after internal storage and distribution. The function of internal storage and distribution is realized by the voD video source station. In the process of internal distribution of the source site, the video data will pass through a large number of server nodes, which are distributed in different regions, different operators’ rooms, different times, and the connectivity is constantly changing. The traditional method is to configure fixed active and standby lines for distribution based on o&M experience. The disadvantage is that the quality of distribution may be seriously affected or even fail when the lines are congested or faulty. PP cloud introduces machine learning algorithm to generate forwarding link scheme in real time and select the best network link to improve distribution speed. This sharing will give a detailed introduction to the optimization scheme architecture, model, algorithm design and practice, as well as future development.


Shen Keyi, Tencent Tiantian Ptu iOS development team leader

Speaker: I graduated from Shanghai Jiaotong University with a master’s degree in 2011 and joined Tencent. I have been engaged in video and image application development. Currently, I am mainly responsible for the development of the iOS version of Tiantian Ptu.

Topic: GPU practice of iOS real-time camera



Sheng Xiaojie, VR technology expert of Youku


Speaker: Sheng Xiaojie, VR technology expert of Youku, graduated from The Institute of Image Communication of Shanghai Jiao Tong University. He joined Alibaba in 2013 and began to be responsible for the algorithm research and architecture construction of youku VR end-to-end vod/live broadcast system in 2016. I am interested in video encoding and decoding, ULTRA high definition video processing, and computer vision.

Topic: Evolution and challenges of UHD VR live broadcasting technology

With the continuous growth of virtual reality content and the gradual popularization of terminal head display devices, under the current limited transmission bandwidth, considering that the resolution of VR content is generally in 4K and above, how to provide users with high fluency and high-definition virtual reality viewing experience has become a technical problem. In addition, how to process and distribute large-scale video data in real time in the scenario of VR live broadcasting also poses great challenges to the algorithms and data processing capabilities of cloud. This share will start with an introduction to the main technical challenges of UHD live VR, and the evolution of solutions proposed by the industry in the face of these challenges. It will also introduce the current progress of virtual reality video coding standards. Finally, the difficulties VR video technology still faces are introduced.


Cui Wenxiu director of Beijing Trimaran Technology Co., LTD

Speaker: More than 13 years of working experience in fortune 500 IT enterprises (Citrix, Symantec), with rich experience in r&d, after-sales, implementation, consulting, pre-sales and sales for enterprise customers. I once led the scheme design, project implementation, operation and maintenance support of information security projects and cloud computing projects of the world’s largest bank, insurance and telecom companies.

Topic: Real-time audio and video technology enables traditional industries

With the continuous development of science and technology, audio and video has gradually become an indispensable part of People’s Daily life because it carries more content and brings better experience. Meanwhile, traditional industries have begun to explore how to fit in with audio and video technology, and have derived explosive products such as online doll catching and live answering questions. This sharing will introduce the development status of real-time audio and video technology, analyze the problems existing in traditional industries and the opportunities and challenges in the future, and illustrate the application prospect of real-time audio and video technology in traditional industries with cases.


Ruitao CAI is in charge of graphics engine of Phase Core Technology

CAI Ruitao, head of Graphics Engine of Xiangxin Technology, born in 1990, master degree, graduated from CAD&CG Laboratory of Zhejiang University. I am mainly engaged in the development of 2D/3D AR graphics engine, and have a deep understanding of intelligent facial beauty, skin beauty, filter effects, 2D/3D stickers, gesture recognition, background segmentation, 3D face reconstruction and facial expression animation synthesis in graphic and image processing.

Topic: Animoji dynamic expression technology implementation


Exquisite gift


LiveVideoStack prepares hot Technology Book: The Definitive Guide to WebRTC

More on the scene:

  • WebRTCon 2018 Conference tickets

  • Exquisite notebook set

  • Multi-function switch plug

  • Mark cup

  • T-shirt

  • tableware

  • slippers

Those who buy VIP tickets will have more opportunities to participate in the lecturer exchange dinner.


[New trend of multimedia development: Beijing Station] countdown 2 weeks


Here, we have prepared for you:

  • “Artificial Intelligence Enhanced real-time Video Super Resolution” Bao Jinlong senior multimedia development engineer

  • Application exploration of Panoramic Live Broadcasting in the Field of Monitoring hou Jinting, Deputy Director of Strategy and Implementation Department of Ricoh Software Research Institute

  • Technology Evolution of Audio Codec Huang Chuanzeng, head of multimedia audio technology in Toutiao Interactive Entertainment Department

  • “Real-time audio and video Technology enables Traditional Industries” Cui Wenxiu, Vice President of Products of Beijing Trimaran Technology Co., LTD

  • Technology Development Trend from audio Replay link wang Fufu, Director of Xiaomi Electronics Acoustics in Beijing

  • Mystery guest