Motion likelihood and proposal modeling in Model-Based Stochastic Tracking

Particle filtering is now established as one of the most popular methods for visual tracking. Within this framework, there are two important considerations. The first one refers to the generic assumption that the observations are temporally independent given the sequence of object states. The second consideration, often made in the literature, uses the transition prior as proposal distribution. Thus, the current observations are not taken into account, requesting the noise process of this prior to be large enough to handle abrupt trajectory changes. As a result, many particles are either wasted in low likelihood regions of the state space, resulting in low sampling efficiency, or more importantly, propagated to distractor regions of the image, resulting in tracking failures. In this paper, we propose to handle both considerations using motion. We first argue that in general observations are conditionally correlated, and propose a new model to account for this correlation allowing for the natural introduction of implicit and/or explicit motion measurements in the likelihood term. Secondly, explicit motion measurements are used to drive the sampling process towards the most likely regions of the state space. Overall, the proposed model allows to handle abrupt motion changes and to filter out visual distractors when tracking objects with generic models based on shape or color distribution. Experimental results obtained on head tracking, using several sequences with moving camera involving large dynamics, and compared against the CONDENSATION algorithm, have demonstrated superior tracking performance of our approach.

Related material