Martina Uray

Learn More
Incremental subspace methods have proven to enable efficient training if large amounts of training data have to be processed or if not all data is available in advance. In this paper we focus on incremental LDA learning which provides good classification results while it assures a compact data representation. In contrast to existing incremental LDA methods(More)
In the paper we propose a novel method for incremental visual learning by combining reconstructive and discriminative subspace methods. This is achieved by embedding LDA learning and classification into the incremental PCA framework. The combined subspace consists of a truncated PCA subspace and a few additional basis vectors that encompass the(More)
In this paper we consider the limitations of Linear Discriminative Analysis (LDA) when applying it for largescale problems. Since LDA was originally developed for two-class problems the obtained transformation is sub-optimal if multiple classes are considered. In fact, the separability between the classes is reduced, which decreases the classification(More)
Enzymes are becoming increasingly important tools for synthesizing and modifying fine and bulk chemicals. The availability of biocatalysts which fulfil the requirements of industrial processes is often limited. Recruiting suited enzymes from natural (e.g. metagenomes) and artificial (e.g. directed evolution) biodiversity is based on screening libraries of(More)
With the introduction of tissue microarrays (TMAs) researchers can investigate gene and protein expression in tissues on a high-throughput scale. TMAs generate a wealth of data calling for extended, high level data management. Enhanced data analysis and systematic data management are required for traceability and reproducibility of experiments and provision(More)
Bringing robustness into subspace methods is very important, for training as well as for recognition. In case of Linear Discriminant Analysis (LDA) the task of robust classification is already solved, therefore, we focus on treating pixel outliers and occlusions in the training stage. More precisely, in this work we consider the task of incremental(More)
Generation of highly active, specific and enantioselective biocatalysts is of increasing interest for industrial processes. That leads to the necessity of reliable high throughput screening methods suitable for the desired enzyme reaction. A new approach is the screening of whole cell arrays where no decomposition of cells and further cleaning steps are(More)
In air traffic management (ATM) all necessary operations (tactical planing, sector configuration, required staffing, runway configuration, routing of approaching aircrafts) rely on accurate measurements and predictions of the current weather situation. An essential basis of information is delivered by weather radar images (WXR), which, unfortunately,(More)
This thesis is focused on Linear Discriminant Analysis (LDA), which is a subspace learning method. LDA is employed for appearance-based object classification. The standard LDA needs all training data to be given in advance in order to construct the subspace. This type of learning is termed batch learning. But in general, not all data is available at the(More)
In this paper an optical system for measuring the amount of dust deposition on commercial bag filters is presented. Dust laden gas is blown onto the filter, the dust remains on the surface and forms a compact layer called filter cake. For academic simulations as well as practical filter design it is desirable to know the thickness of this cake over the(More)