A Toolkit for Gaming Videos
adaptive transcoding and compression services for game media content
automatic video preview thumbnail and highlight reel generation
automatically identify, categorize, label and store game media assets
utilizes Convolutional Neural Networks to rectify video distortion
We provide adaptive transcoding and compression services for game media content, drastically reducing storage and transmission bandwidth requirements. We currently support a range of formats, including H.264, H.265, VP8, VP9, and AV1. Our platform transcodes different compression formats, resolutions, and bitrates and comes prepackaged with time saving presets for the gaming industry.
This service offers automatic highlight reel generation based on accurate event detection, classification and localization as well as frame-by-frame analysis. Automatic thumbnail cover generation is also available, with scenes selected based on recognition, aesthetics and visual quality. In addition, deep learning technology together with signal processing algorithms automatically identify, categorize, label and store game media assets so you can efficiently manage and locate your media assets in a centralized location.
This service leverages powerful CVM processing capabilities and utilizes traditional video processing algorithms together with modern deep learning technologies such as Convolutional Neural Networks to rectify image degradation issues resulting from video compression. It also offers advanced technologies such as super resolution, color enhancement and sharpening to improve the viewing experience.
Highlight reels can be generated by auto-selecting prominent scenes such as shooting and kills from live or recorded streams. These reels can be used for high quality gaming ads and reduces labor costs of manual video selection and editing.
Automatically detects in-game events and generates labels for push notifications. Notifications are pushed in time for higher relevancy and accuracy.
Highlight reels can be automatically generated from live game streams for posting on short video platforms. Factors such as uniqueness, intensity and aesthetics are taken into consideration to select scenes with the best visual representativeness to increase viewership.
Deep neural network is used for automatic asset optimization and self-categorization, enabling one-stop media assets management. This facilitates video searches and recommendations, achieving higher efficiency and reducing costs.
