We propose a video event analysis framework based on object segmentation and tracking, combined with a Hidden Semi-Markov Model (HSMM) that uses state occupancy duration modeling....
Background: Channel current feature extraction methods, using Hidden Markov Models (HMMs) have been designed for tracking individual-molecule conformational changes. This informat...
Stephen Winters-Hilt, Matthew Landry, Mark Akeson,...
The best of Kalman-filter-based frameworks reported in the literature for rigid object tracking work well only if the object motions are smooth (which allows for tight uncertainty ...
— We present an algorithm for learning 3D object models from partial object observations. The input to our algorithm is a sequence of 3D laser range scans. Models learned from th...
Michael Ruhnke, Bastian Steder, Giorgio Grisetti, ...
— This paper makes new contributions in motion detection, object segmentation and trajectory estimation to create a successful object tracking system. A new efficient motion det...