"The program uses an algorithm that draws upon a database of thousands of faces as a reference. It looks at distinguishing points of the face, from the shape of the ears and eyes to hair color, to determine the age. The database expands as more people walk past the camera, allowing the program to make better judgment calls with time," Yamamoto said. "The technology originated in Japan, where it was easier to implement because Japanese physical traits are more uniform. Mr. Yamamoto acknowledged that the U.S. market would be more difficult because of the diverse population."
NEC Using Hair Color, Ear Shape To Help Digital Signage Guess Consumers' Age
Barely a week after Intel and Microsoft announced their own futuristic digital signage plans—including PDA integration and analytics software linked to hidden digital cameras—NEC is talking up its own approach to digitally guessing consumers' age and gender (plus counting them, but that's so 1990s). Takeshi Yamamoto, vice president of NEC America's IT software group, is quoted in the Wall Street Journal saying that the program is pretty good at getting within 10 years of a consumer's age, that the program tracks a person's age and gender but discards the digital footage, and that the data is aggregated.