TU 4: Visual computing
[R. Klein, T. Lippert, M. Reuter]
The field of visual computing includes all disciplines that handle or generate images and geometric models, such as computer graphics, image processing, visualization, computer vision, pattern recognition, human computer interaction and machine learning. With the advent of large-scale simulations and high-throughput data acquisition, visual computing has become essential for quantification, feature extraction, validation and communication. A majority of modern data sources are image or geometry based, geometric representations of biological structures, molecular geometries or visualizations of Quantum Chromodynamics). Many visualization and analysis methods developed in medical image processing transfer directly to other fields, such as physics or chemistry (and vice-versa), where spatial or high-dimensional data needs to be inspected, interpreted, registered, segmented, featurized and quantified. Most recently, for example, Visual Analytics has established itself as a sub-discipline that focuses on supporting analytical reasoning by interactive graphical interfaces that integrate techniques from information and scientific visualization with tools from automated data analysis and, in particular, from statistics and machine learning. In this context, the interdisciplinary TU “Visual Computing” will provide an indispensable interface between scientists, their data, and their computational analyses. It, therefore, bridges across all of CASCADE RUs and links directly with other topical units, most closely with data-intensive modeling, machine learning and network models.
