This thesis is concerned with intention recognition for a humanoid robot and investigates how the challenges of uncertain and incomplete observations, a high degree of detail of the used models, and real-time inference may be addressed by modeling the human rationale as hybrid, dynamic Bayesian networks and performing inference with these models. The key focus lies on the automatic identification of the employed nonlinear stochastic dependencies and the situation-specific inference.

This book gathers revised and extended versions of the best papers presented at the 8th International Joint Conference on Computational Intelligence (IJCCI 2016), which was held in Porto, Portugal from 9 to 11 November 2016. The papers address three main fields of Computational Intelligence, namely: Evolutionary Computation, Fuzzy Computation, and Neural Computation. In addition to highlighting recent advances in these areas, the book offers veteran researchers new and innovative solutions, while also providing a source of information and inspiration for newcomers to the field.

The goal of this work is improving existing and suggesting novel filtering algorithms for nonlinear dynamic state estimation. Nonlinearity is considered in two ways: First, propagation is improved by proposing novel methods for approximating continuous probability distributions by discrete distributions defined on the same continuous domain. Second, nonlinear underlying domains are considered by proposing novel filters that inherently take the underlying geometry of these domains into account.

In robotic beating heart surgery, a remote-controlled robot can be used to carry out the operation while automatically canceling out the heart motion. The surgeon controlling the robot is shown a stabilized view of the heart. First, we consider the use of directional statistics for estimation of the phase of the heartbeat. Second, we deal with reconstruction of a moving and deformable surface. Third, we address the question of obtaining a stabilized image of the heart.

In Networked Control Systems (NCS), components of a control loop are connected by data networks that may introduce time-varying delays and packet losses into the system, which can severly degrade control performance. Hence, this book presents the newly developed S-LQG (Sequence-Based Linear Quadratic Gaussian) controller that combines the sequence-based control method with the well-known LQG approach to stochastic optimal control in order to compensate for the network-induced effects.

A ubiquitous challenge in many technical applications is to estimate an unknown state by means of data that stems from several, often heterogeneous sensor sources. In this book, information is interpreted stochastically, and techniques for the distributed processing of data are derived that minimize the error of estimates about the unknown state. Methods for the reconstruction of dependencies are proposed and novel approaches for the distributed processing of noisy data are developed.

State estimation techniques for centralized, distributed, and decentralized systems are studied. An easy-to-implement state estimation concept is introduced that generalizes and combines basic principles of Kalman filter theory and ellipsoidal calculus. By means of this method, stochastic and set-membership uncertainties can be taken into consideration simultaneously. Different solutions for implementing these estimation algorithms in distributed networked systems are presented.

We discuss theory and application of extended object tracking. This task is challenging as sensor noise prevents a correct association of the measurements to their sources on the object, the shape itself might be unknown a priori, and due to occlusion effects, only parts of the object are visible at a given time. We propose an approach to track the parameters of arbitrary objects, which provides new solutions to the above challenges, and marks a significant advance to the state of the art.

The focus of this work is a generic, intraoperative and image-free planning and execution application for arbitrary orthopedic interventions using a novel handheld robotic device and optical see-through glasses (AR). This medical CAD application enables the surgeon to intraoperatively plan the intervention directly on the patient's bone. The glasses and all the other instruments are accurately calibrated using new techniques. Several interventions show the effectiveness of this approach.

This work is concerned with the simultaneous tracking and shape estimation of a mobile extended object based on noisy sensor measurements. Novel methods are developed for coping with the following two main challenges: i) The computational complexity due to the nonlinearity and high-dimensionality of the problem, and ii) the lack of statistical knowledge about possible measurement sources on the extended object.