Ing. Karel Fliegel, Ph.D.

There is a current need for efficient image compression techniques and related metrics for image quality evaluation to be addressed in the area of emerging image modalities, including light field, point cloud, and digital holography. The thesis will focus on the development of efficient image compression schemes for these modalities and their robust performance evaluation.

Users of the latest information technology and communication networks produce and share the ever-increasing amount of visual content. With the development of artificial neural network models, highly parallel computing systems with increasing computing power, and the availability of extensive databases of training and testing image data, the trend of machine learning methods for the efficient representation of image content can be traced. The dissertation will focus on the development of methods for the efficient representation of image data as well as on related tools for performance evaluation of the proposed algorithms based on machine learning techniques.