...
"Mining" Video For Profits
The retail business can be tough. The wrong display can turn people off, customers walk away if the checkout line is too long, and shoplifters eat away at slim profit margins. Advanced Interfaces Inc. and its VideoMining division are out to change all that.
“The lifetime value of a customer is very high,” says Jeff Hershey, vice president of business development for VideoMining, which is AI’s first commercial division. Retailers can use Video-Mining technology to attract customers and give them experiences that bring them back again and again, he says.
Jeff Hershey showcases Advanced Interface video-imaging software at the National Retail Federation’s 94th Annual Convention and Expo held at the Javits Center in Manhattan in January 2005.
Most retailers already have video cameras in place as a security measure. VideoMining takes these cameras to the next level by analyzing or “mining” videos, extracting behavioral characteristics and sending alerts when a response is needed. VideoMining products automatically recognize more than 100 human behaviors and characteristics. For example, Hershey explains, losses can be prevented by alerting security when suspicious behaviors often associated with shoplifting are spotted.
Retailers can also ask the software to search for specific occurrences over the course of several days. “Store owners don’t have time to look through hours of video to find all the instances where a person waited at the counter for more than 30 seconds,” Hershey says. VideoMining can pull up all those instances, allowing the retailer to quickly review footage, find out why customers were waiting too long, and then make changes in processes or employees. “It allows retailers to monitor and manage all the customer touchpoints,” according to Hershey.
Other retail uses for VideoMining technology include:
• Customer traffic: for example, determining whether more batteries are being sold from displays near the photo lab or from displays near the checkout.
• Customer segmentation: using gender-recognition software to find out whether men or women are buying more at different times of day.
• Media measurement: measuring and targeting on-site advertising based on demographics, average time spent in a certain area, etc.
“ The technology is quite revolutionary,” says Tony Warren, director of Penn State’s Farrell Center for Corporate Innovation & Entrepreneurship and an advisor to Advanced Interfaces. “Now the retail sector is beginning to realize the power of what they’ve got.”
Outside retail, AI is working on putting its technology to use in several other sectors
• Large-area surveillance: using a database of suspicious behaviors to automatically monitor individuals in airports and other public spaces as they travel from one video monitor to another, alerting security personnel when needed.
• Remote care of the elderly: monitoring a large number of nursing-home residents simultaneously and alerting staff when a fall or other mishap occurs.
• Assistive communication: allowing personalized computer access for people with severe communication difficulties (such as paralysis) by filtering out unintentional movements and providing multiple ways of controlling functions to reduce stress and fatigue.
Founded by Penn State Professor Rajeev Sharma, Advanced Interface already has 23 patents pending for its technology, and more are on the way, Hershey says. The company employs about 30 people, many of them Penn State graduates, in its offices and lab in The Towers on South Allen Street. Much of AI’s research is funded through government grants, followed by incubation in commercial applications.
“ I think the next 12 to 18 months will see an enormous difference in this company,” Warren says.
...
