Articulated object tracking has now become a very active research area in the field of computer vision. One of its applications, i.e. human tracking, is used in a variety of domains, such as security surveillance, human computer interface, gait analysis,...The problem is also of interest from the theoretical point of view. Some of its challenges include, for example, the high dimensionality of state spaces, self-occlusions, kinematic ambiguities or singularities, making it hard to solve and hence, attractive for the tracking community. Particle Filter (PF) has been shown to be an effective method for solving visual tracking problems. This is due to its ability to deal with non-linear, non-Gaussian and multimodal distributions encountered in such problems. The key idea of particle filter is to approximate the posterior distribution of the target object state by a set of weighted samples. These samples evolve using a proposal distribution and their weights are updated by involving new observations. Unfortunately, in high dimensional problems, such as articulated object tracking problems, the number of samples required for approximating the target distribution can be prohibitively large since it grows exponentially with the number of dimensions (e.g., the number of parts of the object), making the particle filter impractical. To reduce the complexity of tracking algorithms in such problems, various methods have been proposed. One family of approaches that has attracted many researchers is based on the decomposition of the state space into smaller dimensional sub-spaces where tracking can be achieved using classical methods. This results in tracking algorithms that are linear instead of exponential in the number of parts of the object.