Media Summary: IMRL: An Improved Inertial-Aided KLT Feature Tracker Clustering people based on their motion in real time. Chenge Li, New York University.
IMRL: An Improved Inertial-Aided KLT Feature Tracker Clustering people based on their motion in real time. Chenge Li, New York University.
Filmed and coded by Shih-En Wei.
IMRL: An Improved Inertial-Aided KLT Feature Tracker
Output from
KLT tracking
Pyramidal Kanade-Lucas-Tomassi
features klt all
Detection and
Sequence from the DroneCrowd dataset.
Clustering people based on their motion in real time. Chenge Li, New York University.
KLT feature tracking
This video uses Birchfield's
These tracks were detected by the
People Clustering using