Mobile phones are becoming inseparable from humans and are replacing
traditional cameras.
TagSense attempts to leverage this trend to automatically tag pictures.
The core idea of TagSense is to gather sensor
information from nearby participating phones at the time of picture click.
It then processes the sensor data to tag the picture
with people and context.
TagSense approach to tagging people is complementary to traditional face recognition
based techniques.
Our paper
and poster
on TagSense are presented at MobiSys
2011.
Complete presentation is available in two parts:
part1 and
part2.
Excerpts of paper and presentation are shown
below.
TagSense received coverage by the popular media. Here is a representative list of links to articles about TagSense.
Some of these are picked up by University of South
Carolina Media Relations and inadvertently by the other USC.
