A human visual standard-based game video processing platform combining deep learning with signal processing technology for the best viewing experience
Game Video Processor (GVP) fixes and enhances the quality of game videos. Specifically, it rectifies video distortions caused by excessive compression through deep learning and improves video quality through color enhancement, super-resolution, and edge sharpening for an optimized viewing experience.
GVP automatically analyzes the distortion intensity in a video to call the corresponding artifact removal model. It selects the most appropriate video quality enhancement method based on video scene.
GVP determines video quality based on human visual characteristics to avoid quality degradation caused by excessive enhancement.
GVP is pay-as-you-go and can be run in either CPU or GPU, helping you balance processing speed and costs.
GVP leverages deep learning technology to rectify distortion issues resulting from video compression, such as blocking and ringing artifacts.
GVP can adaptively enhance video color saturation and dynamic range of brightness according to video content to make videos more attractive.
By sharpening the edges and textures of a video and utilizing the super-resolution algorithm based on deep learning, GVP can improve the resolution and clarity of the video to display more details.
GVP can restore and enhance the quality of gaming videos for an improved viewing experience, making it ideal for game VOD platforms.
Noises caused by compression greatly compromise the accuracy of video analysis. By restoring video quality, GVP helps address this issue for video analysis applications such as categorizing and tagging.
In order to save storage space and reduce transfer bandwidth, videos usually need to be transcoded in the cloud. GVP features video enhancement that can optimize the video output’s playback quality while lowering bitrate during video transcoding.