Tracking-as-Recognition for Articulated Full-Body Human Motion Analysis
MetadataShow full item record
This paper addresses the problem of markerless tracking of a human in full 3D with a high-dimensional (29D) body model. Most work in this area has been focused on achieving accurate tracking in order to replace marker-based motion capture, but do so at the cost of relying on relatively clean observing conditions. This paper takes a different perspective, proposing a body-tracking model that is explicitly designed to handle real-world conditions such as occlusions by scene objects, failure recovery, long-term tracking, auto-initialisation, generalisation to different people and integration with action recognition. To achieve these goals, an action's motions are modelled with a variant of the hierarchical hidden Markov model. The model is quantitatively evaluated with several tests, including comparison to the annealed particle filter, tracking different people and tracking with a reduced resolution and frame rate.
Showing items related by title, author, creator and subject.
Mallick, M.; Rubin, S.; Vo, Ba-Ngu (2013)Space object (satellite or space-debris) tracking (SOT) has not received much attention in the Information Fusion community, although the first Fusion conference was held in 1998. A special session on SOT was organized ...
Shahin, Mohamed (2008)Due to the daily congestion of highways, railways have become the most popular means of public transportation, which increased the demand for heavier and faster trains while keeping the cost of track maintenance at its ...
Xia, Jianhong (Cecilia); Basic, F.; Arrowsmith, C. (2005)The aim of this paper is to identify suitable tracking techniques for the spatio-temporal movement behaviour of tourists in a natural environment for various applications. Several techniques at both a micro and macro scale ...