Wangsu’s intelligent cloud video platform enables AI and big data technologies in all video service processes and implements product intelligence based on operational intelligence. With the support of content delivery networks worldwide, it realizes product intelligence across the entire video chain, from video production to processing, delivery and playback, and provides full support for VOD (short and long videos), live streaming and other services.
The VOD platform is proprietarily developed by Wangsu. Through next-generation intelligent scheduling system (NGB), file prefetch, file push, multiple file uploading methods, exclusive cache and other functions, and in connection with video drag and play, video screenshot, multi-screen play resumption, package conversion and transcoding as well as other functions, it enables device content adaptation and fully meets the needs of different types of VOD businesses. As for short videos, the platform is integrated with video AI, P2P and big data analysis technologies to conduct intelligent labeling classification of videos and perform content audits as well as provide dynamic special effect covers. Meanwhile, in relation to the player SDK, it supports the instant loading of videos.
Intelligent labeling: Make label-based classification of contents, and improve content search and recommendation precision.
Subtitle generation: Intelligently recognize audio contents, make voice transcriptions, and add subtitles to videos.
Content auditing: making multidimensional examination of image, voice and text contents, and perform efficient authentication of pornographic, violence, politically sensitive and self-defined contents.
Adaptive coding and decoding: Full-link support for H.265 coding and decoding, achieve lower code rates under the same video quality and clearer images with the same bit rate, in addition to supporting H.264 and H.265 adaptive matching.
HLS advert inter-cut: Insert adverts at designated positions, and help customers to generate operating revenue.
Movie trial screening: Offer responsive trials to non-VIP users, boost users’ willingness to pay and help customers to promote resources.
AI video enhancement: Through deep learning-based AI video enhancement technology, intelligently enhance image effects and definition; remove burrs, artifacts, noises and mosaics, to improve the viewing experience.
Resolution reconstruction: Deep learning-based video ultra-resolution technology breaks resolution constraints, automatically restore video picture quality thus enabling low-resolution videos to deliver a high-resolution viewing experience.
Intelligently controlled HD: Business scenario-based intelligent matching, with lossless bit reduction, picture restoration and other technologies to achieve high-definition viewing under narrow bandwidth, streaming of old films no longer suffers from poor quality.
VR on-demand transmission: Through AI and VR-based business scenario training, achieves on-demand transmission, optimize video quality in critical zones, reduce total bit rate and ensure user experience.
Multidimensional copyright protection: Support conventional referrer anti-theft chain, IP black and white lists, UA anti-hotlinking, timestamp anti-hotlinking, back-to-the-source authentication and other copyright protection methods. In addition, Wangsu has launched exclusive network-wide central authentication anti-hotlinking technology which enables multiple protection in combination with link anti-hijacking policy, video DRM encryption technology and instant deletion, one-click streaming ban and other functions, to offer greater video copyright security and safeguards, across the video, delivery and playback levels.
P2P fusion delivery: CDN+P2P multi-platform cross-device delivery mode, which ensures low-delay, high-sharing, high-fluency and high-stability, maximizes the value of idle resources and reduces delivery cost. Connecting the mobile end (iOS/Android), the service end (CDN/Html5) to achieve overall efficiency optimization
Terabit-level bandwidth bursts: Can carry out millions of concurrent user requests, and support Terabit-level bandwidth bursts.
The VOD platform has four major systems: intelligent operational diagnosis, intelligent scheduling, big data analysis and differentiated operations. Based on user IP, content, network, health value and other determinations and in relation to big data analysis technology, it achieves automatic diagnosis and analysis of full-link problems and provides differentiated operational solutions at the business level.