Your muploads album just got ten times better.
With a new mobile application called TagSense developed at Duke and University of South Carolina, smartphone users will be able to automatically tag people, location, surroundings and actions occurring in a photograph.
TagSense is much more sophisticated than basic facial recognition in the ways that it incorporates many more details. The application recognizes human behavior with an accelerometer based on motion signatures, complementary compass directions and identifying moving subjects who are running, dancing or playing a game.
When someone takes a picture of his friends, his phone will quickly and wirelessly connect to his friend’s phones and instruct them to activate their sensors, explained Xuan Bao, a Ph.D. student in computer science at Duke and developer of this application.
“The sensors sense the ‘moment,’ summarize the measured data and communicate them back to [his] phone," Bao wrote in an email. "[His] phone processes this data to identify which of the individuals are in the picture, their activities and other contextual tags about the occasion. These tags are systematically organized into a ‘when-where-who-what’ format, ultimately creating an automatic description for the picture.”
Bao envisions users who like to take photos on their smartphones as the target market segment for this application, which will make it easier for them to search and share pictures.
Rather than searching an entire photo library for a specific picture of a friend running in a race, for example, the smartphone user could search for photos that tag his friend, someone running and a race. TagSense makes this process much quicker and provides a more intuitive and comprehensive way of organizing the photos.
TagSense was unveiled at the ninth Association for Computing Machinery's International Conference on Mobile Systems, Applications and Services (MobiSys) held in Washington, D.C. on Tuesday. The application will probably not hit the App Store for another few years, however—they are currently still working on the prototype, Bao said.
“The big vision is to fully utilize the rich sensor set available on advanced mobile phones to enable a new generation of context-free applications,” Bao said. “Imagine an array of exciting applications [that] can be supported when the information of activities, participants and environment can be learned by the applications themselves.”
Get The Chronicle straight to your inbox
Signup for our weekly newsletter. Cancel at any time.