No abstract

This paper introduces a new visual tracking technique combining particle filtering and Dynamic Bayesian Networks. The particle filter is utilized to robustly track an object in a video sequence and gain sets of descriptive object features. Dynamic Bayesian Networks use feature sequences to determine different motion patterns. A Graphical Model is introduced, which combines particle filter based tracking with Dynamic Bayesian Network-based classification. This unified framework allows for enhancing the tracking by adapting the dynamical model of the tracking process according to the classification results obtained from the Dynamic Bayesian Network. Therefore, the tracking step and classification step form a closed trackingclassification-tracking loop. In the first layer of the Graphical Model a particle filter is set up, whereas the second layer builds up the dynamical model of the particle filter based on the classification process of the Dynamic Bayesian Network.

In this article a novel approach towards managing unmanned aerial vehicle missions based on the human operator's intent is introduced. The system architecture comprises two artificial cognitive agents, one conducting semi-autonomous mission execution aboard the vehicle and the other providing assistance to the operator. The assistant system supports the operator in expressing his intent to the technical system by adapting to his mental state and cognitive processes. To enable the assistant system to gain insight into the operator's state of mind we created a model of his tasks, knowledge, and cognitive processes while managing the mission. Based on this model the assistant system infers the operator's cognitive state and interacts adaptively. With that, human error is minimized and the operator's mission management process is improved regarding workload and mission success. The overall system has been used to conduct a successful campaign of preliminary human-machine and real-flight UAV experiments.

This paper introduces our research platform for enabling a multimodal Human-Robot Interaction scenario as well as our research vision: approaching problems in a holistic way to realize this scenario. However, in this paper the main focus is laid on the image processing domain, where our vision has been realized by combining particle tracking and Dynamic Bayesian Network classification in a unified Graphical Model. This combination allows for enhancing the tracking process by an adaptive motion model realized via a Dynamic Bayesian Network modeling several motion classes. The Graphical Model provides a direct integration of the classification step in the tracking process. First promising results show the potential of the approach.