blinkx CORE is the world’s most advanced video engine.
Based on technology that was developed over 12 years at Cambridge University and is protected by more than 111 patents, blinkx CORE solves the challenges inherent in processing, managing and monetizing rich media.
The CORE engine automates traditionally expensive and time-consuming manual tasks including processing, analysis and tagging, and exposes the full potential of audio and video content.
From locating and processing online and offline content, and indexing it, to its presentation, monetization and delivery, it unlocks the true value of rich media and creates new opportunities for monetization.
To achieve this, blinkx CORE first distills video into its basic building blocks and then processes each component in the method most effective to the medium: audio is processed with speech recognition technology; images are understood using visual analysis technology; and metatags are indexed and extracted by distributed visual spiders.
Once it has collected the most comprehensive, minutely accurate data possible on the video, the CORE engine transforms the raw data into useful information by recognizing ideas and themes within the video.
Using a unique combination of Shannon's Information Theory and Bayesian Inference, blinkx CORE applies a probabilistic pattern matching process to the problem of concept matching. This enables computers to infer meaning within pieces of video and audio content, and reason upon the nature of that meaning in an unstructured way, performing the kinds of organizational, sorting and discovery tasks that today are the domain of manual, human processes. Portions of the blinkx CORE engine include technology used under license from Autonomy.
This deep, granular understanding of rich media enables blinkx to process, present, monetize and deliver video and audio content in unique ways, and to capitalize on the true potential of today’s Four Screen world.