Wangsu’sintelligent cloud video platform enables AI and big data technologies in allvideo service processes and implements product intelligence based onoperational intelligence. With the support of content delivery networksworldwide, it realizes product intelligence across the entire video chain, fromvideo production to processing, delivery and playback, and provides fullsupport for VOD (short and long videos), live streaming and other services.
Proprietarilydeveloped by Wangsu, the live video streaming platform supports multipleprotocols, including RTMP, HTTP FLV, HTTP TS, HLS and HDS, to meetmulti-device, multi-platform business needs. It offers an integrated solutioncovering video stream push to video streaming processing and end-pointplayback. With full-link intelligent scheduling, real-time big data analysis,multidimensional automatic operational diagnosis and in relation to AItechnology, it is an intelligent live streaming platform covering from thefirst to the last mile of video streaming. The platform provides multiple SDKand API interfaces to facilitate quick, stable and secure access to live streamingservices, enabling users truly enjoying low-delay and high-fluency audiovisuallive streaming experience.
Streampusher: Provides interactive live streaming at the mobile end (including iOSand Android) with data collection, coding, transmission as well as other basicfunctions. It possesses auxiliary functions including beautification and weaknetwork policies.
Videoplayer: Provides the function of playing video streams or video files, coveringall mainstream audiovisual formats.
Miclink: By embedding SDK into devices, supporting multi-person mic link,intelligent selection of optimal mic link nodes, and provide users with miclink services with millisecond-level delay
Cloudstreaming direction: a “cloud”-based directing station which supports multiplefunctions on the cloud, including traditional directing switchovers, synthesis,barrage subtitles, watermarks and others.
Content auditing: Perform multidimensional auditing of images, voices, texts and other contents, and efficiently identifies pornographic, violence, politically sensitive and self-defined contents.
Adaptivecoding and decoding: Full-link support for H.265 coding and decoding, achievelower code rates under the same video quality and clearer images with the samebit rate, in addition to supporting H264 and H265 adaptive matching.
Resolution reconstruction: Deep learning-based video ultra-resolution technology which can make in-depth reconstruction of video images on playback devices, break resolution constraints, achieve 2x and 3x video over resolution thus enabling low-resolution videos to deliver a high-resolution viewing experience.
Intelligently controlled HD:Technologieslike business scenario-based intelligent recognition, lossless bit reduction,picture restoration and etc. achieve high-definition viewing at low bit rates.Streaming of old films no longer suffers from poor quality.
WNHD（Weak network HD）: By embedding SDK at the mobile end,establishing a bilateral interaction mechanism and improve stream push qualityat the last mile in relation to intelligent network state recognition, privatetransmission protocol and other technologies.
VRlive streaming: Through machine learning and VR business scenario training,optimize model algorithms, dynamically adjust effective pixels and imagequality under primary and non-primary viewing angles. Save transmissionbandwidth, improve transmission efficiency, and address the difficult issues ofhigh VR video costs and high network environment requirements without affectinguser experience.
Multidimensionalcopyright protection: Support conventional referrer anti-theft chain, IP blackand white lists, UA anti-hotlinking, timestamp anti-hotlinking,back-to-the-source authentication and other copyright protection methods. Inaddition, Wangsu has launched exclusive network-wide central authenticationanti-hotlinking technology which enables multiple protection in combinationwith link anti-hijacking policy, video DRM encryption technology and instantdeletion, one-click streaming ban and other functions, to offer greater videocopyright security and safeguards.
P2Pfusion delivery: CDN+P2P multi-platform cross-device delivery mode, whichenables low-delay, high-sharing, high-fluency and high-stability businessfeatures, maximizes the value of idle resources and reduces delivery cost.Connecting the mobile end (iOS/Android), the PC end (Flash/HTML5) and theservice end to achieve overall efficiency optimization.
Terabit-levelbandwidth bursts: Can carry out tens of millions of concurrent user requests,and support Terabit-level bandwidth bursts.
The live video streaming platform consists of four major systems: intelligentoperational diagnosis, intelligent scheduling, big data analysis anddifferentiated business operations. Incombination with big data analysis technology, it supports the diagnosis offull-link automation issues. Based on user IP, content, network, health valueand multiple data, it achieves instant intelligent scheduling. It is anintegrated intelligent data analysis platform that provides data collection,mining, analysis and other multiple operations. Based on business andconfiguration differences, it intelligently switches between operationalpolicies. Intelligent operations can effectively safeguard the live videostreaming platform in providing stable and efficient services.