Video-Inform's Video Analytics ExaminedBy John Honovich, Published on Sep 01, 2009
A new video analytic manufacturer launched in 2008 aims to provide analytics that can be trained to identify sunglasses, scooters, uniforms, backpacks and more.
Video Inform markets a solution they call an "Artificial Profiler" that they claim can "can uniquely distinguish and categorize different objects much like humans do cognitively."
Essentially, the system can be trained manually to look for objects of similar size, shape, color, etc. (like sunglasses, scooters, uniforms, etc.)
In a presentation at ISC West [link no longer available], Video Inform's CTO explained their technology (presentation slides with details [link no longer available]).
From a marketing perspective, their product is similar in concept or potential to BRS Labs. However, unlike BRS Labs, Video Inform is presenting significantly greater technical details. Also, their founders [link no longer available] were previously at Vigilant [link no longer available], providing some track record within the industry.
Of course, critical questions and issues remain with this class of technology including (1) complexity of on-site setup/tuning, (2) integration options with existing VMS systems and (3) cost/quantity of computers needed to analyze.
[Update 2012: We have heard nothing about or from this company in the past 3 years.]