C. Smartphone Application
The application developed on the smartphone side has
two major roles. First, it performs food type recognition
by implementing the main algorithms detailed in the next
section. Second, it serves as a data manager and provides
an interface to the user. Fig. 3 gives some screenshots of
the application. To use our system, a user simply wear and
power on AutoDietary, and start the application. When the
user starts to eat, the system will perform food type recognition
and store the detailed data into a database. The user can not
only check the detailed records (as shown in Fig. 3(b)), but
also go through the suggestions on healthier eating habits
which are obtained by analyzing the data collected. Eating
guidance currently includes: a) more regular and balanced
diets, b) alerts on abnormal chewing speed, c) suggestions
on hydration intake, d) alerts on excessive snacking in a day,
and e) suggested intervals between meals. Based on the key
information and main framework provided by AutoDietary,
developers can further expand the application with new
features on personal health management.