Evolving Teams of Predictors with Linear Genetic Programming

This paper applies the evolution of GP teams to different classification and regression problems and compares different methods for combining the outputs of the team programs. These include hybrid approaches where (1) a neural network is used to optimize the weights of programs in a team for a common decision and (2) a realnumbered vector (the representation of evolution strategies) of weights is evolved with each term in parallel. The cooperative team approach results in an improved training and generalization performance compared to the standard GP method. The higher computational overhead of team evolution is counteracted by using a fast variant of linear GP.     

Interactive volume exploration for feature detection and quantification in industrial CT data

This paper presents a novel method for interactive exploration of industrial CT volumes such as cast metal parts, with the goal of interactively detecting, classifying, and quantifying features using a visualization-driven approach. The standard approach for defect detection builds on region growing, which requires manually tuning parameters such as target ranges for density and size, variance, as well as the specification of seed points. If the results are not satisfactory, region growing must be performed again with different parameters. In contrast, our method allows interactive exploration of the parameter space, completely separated from region growing in an unattended pre-processing stage. The pre-computed feature volume tracks a feature size curve for each voxel over time, which is identified with the main region growing parameter such as variance. A novel 3D transfer function domain over (density, feature size, time) allows for interactive exploration of feature classes. Features and feature size curves can also be explored individually, which helps with transfer function specification and allows coloring individual features and disabling features resulting from CT artifacts. Based on the classification obtained through exploration, the classified features can be quantified immediately.     

3D Video Billboard Clouds

3D video billboard clouds reconstruct and represent a dynamic three-dimensional scene using displacement-mapped billboards. They consist of geometric proxy planes augmented with detailed displacement maps and combine the generality of geometry-based 3D video with the regularization properties of image-based 3D video. 3D video billboards are an image-based representation placed in the disparity space of the acquisition cameras and thus provide a regular sampling of the scene with a uniform error model. We propose a general geometry filtering framework which generates time-coherent models and removes reconstruction and quantization noise as well as calibration errors. This replaces the complex and time-consuming sub-pixel matching process in stereo reconstruction with a bilateral filter. Rendering is performed using a GPU-accelerated algorithm which generates consistent view-dependent geometry and textures for each individual frame. In addition, we present a semi-automatic approach for modeling dynamic three-dimensional scenes with a set of multiple 3D video billboards clouds.     

Local statistics of retinal optic flow for self-motion through natural sceneries

Image analysis in the visual system is well adapted to the statistics of natural scenes. Investigations of natural image statistics have so far mainly focused on static features. The present study is dedicated to the measurement and the analysis of the statistics of optic flow generated on the retina during locomotion through natural environments. Natural locomotion includes bouncing and swaying of the head and eye movement reflexes that stabilize gaze onto interesting objects in the scene while walking. We investigate the dependencies of the local statistics of optic flow on the depth structure of the natural environment and on the ego-motion parameters. To measure these dependencies we estimate the mutual information between correlated data sets. We analyze the results with respect to the variation of the dependencies over the visual field, since the visual motions in the optic flow vary depending on visual field position. We find that retinal flow direction and retinal speed show only minor statistical interdependencies. Retinal speed is statistically tightly connected to the depth structure of the scene. Retinal flow direction is statistically mostly driven by the relation between the direction of gaze and the direction of ego-motion. These dependencies differ at different visual field positions such that certain areas of the visual field provide more information about ego-motion and other areas provide more information about depth. The statistical properties of natural optic flow may be used to tune the performance of artificial vision systems based on human imitating behavior, and may be useful for analyzing properties of natural vision systems.     

Low-cost telepresence for collaborative virtual environments

We present a novel low-cost method for visual communication and telepresence in a CAVEtrade-like environment, relying on 2D stereo-based video avatars. The system combines a selection of proven efficient algorithms and approximations in a unique way, resulting in a convincing stereoscopic real-time representation of a remote user acquired in a spatially immersive display. The system was designed to extend existing projection systems with acquisition capabilities requiring minimal hardware modifications and cost. The system uses infrared-based image segmentation to enable concurrent acquisition and projection in an immersive environment without a static background. The system consists of two color cameras and two additional b/w cameras used for segmentation in the near-IR spectrum. There is no need for special optics as the mask and color image are merged using image-warping based on a depth estimation. The resulting stereo image stream is compressed, streamed across a network, and displayed as a frame-sequential stereo texture on a billboard in the remote virtual environment     

Interfaces as functors, programs as coalgebras—A final coalgebra theorem in intensional type theory

In [P. Hancock, A. Setzer, Interactive programs in dependent type theory, in: P. Clote, H. Schwichtenberg (Eds.), Proc. 14th Annu. Conf. of EACSL, CSL’00, Fischbau, Germany, 21–26 August 2000, Vol. 1862, Springer, Berlin, 2000, pp. 317–331, URL citeseer.ist.psu.edu/article/hancock00interactive.html; P. Hancock, A. Setzer, Interactive programs and weakly final coalgebras in dependent type theory, in: L. Crosilla, P. Schuster (Eds.), From Sets and Types to Topology and Analysis. Towards Practicable Foundations for Constructive Mathematics, Oxford Logic Guides, Clarendon Press, 2005, URL www.cs.swan.ac.uk/csetzer/] Hancock and Setzer introduced rules to extend Martin-Löf's type theory in order to represent interactive programming. The rules essentially reflect the existence of weakly final coalgebras for a general form of polynomial functor. The standard rules of dependent type theory allow the definition of inductive types, which correspond to initial algebras. Coalgebraic types are not represented in a direct way. In this article we show the existence of final coalgebras in intensional type theory for these kind of functors, where we require uniqueness of identity proofs () for the set of states and the set of commands which determine the functor. We obtain the result by identifying programs which have essentially the same behaviour, viz. are bisimular. This proves the rules of Setzer and Hancock admissible in ordinary type theory, if we replace definitional equality by bisimulation. All proofs [M. Michelbrink, Verifications of final coalgebra theorem in: Interfaces as Functors, Programs as Coalgebras—A Final Coalgebra Theorem in Intensional Type Theory, 2005, URL www.cs.swan.ac.uk/csmichel/] are verified in the theorem prover agda [C. Coquand, Agda, Internet, URL www.cs.chalmers.se/catarina/agda/; K. Peterson, A programming system for type theory, Technical Report, S-412 96, Chalmers University of Technology, Göteborg, 1982], which is based on intensional Martin-Löf type theory.     

Programmable logic construction kits for hyper-real-time neuronal modeling

Programmable logic designs are presented that achieve exact integration of leaky integrate-and-fire soma and dynamical synapse neuronal models and incorporate spike-time dependent plasticity and axonal delays. Highly accurate numerical performance has been achieved by modifying simpler forward-Euler-based circuitry requiring minimal circuit allocation, which, as we show, behaves equivalently to exact integration. These designs have been implemented and simulated at the behavioral and physical device levels, demonstrating close agreement with both numerical and analytical results. By exploiting finely grained parallelism and single clock cycle numerical iteration, these designs achieve simulation speeds at least five orders of magnitude faster than the nervous system, termed here hyper-real-time operation, when deployed on commercially available field-programmable gate array (FPGA) devices. Taken together, our designs form a programmable logic construction kit of commonly used neuronal model elements that supports the building of large and complex architectures of spiking neuron networks for real-time neuromorphic implementation, neurophysiological interfacing, or efficient parameter space investigations.     

Analyzing vortex breakdown flow structures by assignment of colors to tensor invariants

Topological methods are often used to describe flow structures in fluid dynamics and topological flow field analysis usually relies on the invariants of the associated tensor fields. A visual impression of the local properties of tensor fields is often complex and the search of a suitable technique for achieving this is an ongoing topic in visualization. This paper introduces and assesses a method of representing the topological properties of tensor fields and their respective flow patterns with the use of colors. First, a tensor norm is introduced, which preserves the properties of the tensor and assigns the tensor invariants to values of the RGB color space. Secondly, the RGB colors of the tensor invariants are transferred to corresponding hue values as an alternative color representation. The vectorial tensor invariants field is reduced to a scalar hue field and visualization of iso-surfaces of this hue value field allows us to identify locations with equivalent flow topology. Additionally highlighting by the maximum of the eigenvalue difference field reflects the magnitude of the structural change of the flow. The method is applied on a vortex breakdown flow structure inside a cylinder with a rotating lid.