News & Press Releases


MoreIdeas is proud to announce funding by the Canada Media Fund for VideoPixx™ as part of their support for innovative digital media projects.


MoreIdeas annouces VideoPixx™ for video frame classification!

 

Technology

MoreIdeas’ technology platform is interdisciplinary, combining the fields of computer vision, learning theory, visual psychophysics, and computational graphics.

The goal of computer vision is to understand how people interpret, or “see” the content and meaning of images and, in turn, build that functionality into a computer –akin to machine learning, but focused on visual perception. MoreIdeas develops learning algorithms and computer vision techniques to automatically extract information about objects on images (facial feature detection, identification and localization) and automatically infer object properties (subject profile inference and classification).

Visual psychophysics is a branch of psychology that measures how people react to perceptual stimuli. It is incorporated into MoreIdeas learning algorithms to guide the results of its computer vision applications. Providing a framework to systemize the concept of visual appearance and accurately predict how people will perceive a change in the user's appearance in a photo. MoreIdeas builds this into image libraries designed to include emotions, attributes, styles (including avatars) and behavioral references.

Computer graphics technologies are used to allow the user to interactively redesign their appearance using the computed visual information and image libraries.

With this as a foundation, MoreIdeas has developed technology platforms to infer semantic and structural information from images. So far, two proprietary technologies: PixxMe and MetaPixx™, to infer and leverage image content on the Internet.