Latest SCI publications
Research project (§ 26 & § 27)
Duration : 2022-05-01 - 2024-01-03
The progress of statistical machine learning methods has made AI increasingly successful. Deep learning exceeds human performance even in the medical domain. However, their full potential is limited by their difficulty to generate underlying explanatory structures, hence they lack an explicit declarative knowledge representation. A motivation for this project are rising legal and privacy issues – to understand and retrace machine decision processes. Transparent algorithms could appropriately enhance trust of medical professionals, thereby raising acceptance AI solutions generally. This project will provide important contributions to the international research community in the following ways: 1) evidence in various methods of explainability, patterns of explainability, and explainability measurements. Based on empirical studies (“How do humans explain ?”) we will develop a library of explanatory patterns and a novel grammar how these can be combined. Finally, we will define criteria/benchmarks for explainability and provide answers to the question “What is a good explanation?”. 2) Principles to measure effectiveness of explainability and explainability guidelines and 3) Mapping human understanding with machine explanations and deploying an open explanatory framework along with a set of benchmarks and open data to stimulate and inspire further research among the international AI/machine learning community. All outcomes of this project will be made openly available to the international research community.
As a result of global warming increased exceptional floods and extreme heavy precipitation events take place. So the risk of remobilization of deposits increases. Subsequently radioactive heavily contaminated sediments can be mobilized. At LLC-Laboratory Arsenal radioactivity of the danube compartiments: water (dissolved radionuclides), suspended matter and sediment are continuously monitored based on monthly composite samples and event-related samples during floods since 1984. This is a unique Central European radioecological long time series of measurements. The continuation of this sampling and data collection is of great importance to meet future challenges in radiation protection with regard to potential large-scale environmental contamination.
Research project (§ 26 & § 27)
Duration : 2022-11-01 - 2025-10-31
The forest inventory provides information and data as a basis for decision-making in forest planning with regard to the implementation of operational interventions and the achievement of economic objectives. In addition to traditional surveying methods, forest inventories with laser scanners have recently become increasingly important. While terrestrial laser scanners (TLS) have been in use for some time, the first experiences and research results on forest inventories with mobile person-carried laser scanners (PLS) have recently become available. In the last six years, automatic algorithms for the evaluation of 3D laser data have been developed at the BOKU Institute of Forest Growth, thus laying the foundation for a modern sensor-based forest inventory. The measurement process with PLS takes about 10 min for a sample circle with 20 m radius. In a series of publications, a large number of reference measurements have shown that tree cover rates of over 97% can be achieved. The mean deviation (bias) of the BHD measurement is 0 - 0.7 cm, with an average error (RMSE) of 1.5 to 3 cm. Tree heights can be automatically determined with a mean deviation of0 to 0.2 m and an average error of 1.5 to 2.1 m. With the portable laser scanner, precise measurement values can now be determined in three-dimensional space in a short time, and the new laser-based forest inventory procedure is thus on the verge of general practical applicability. Together with the Austrian Federal Forests, the new laser-based digital forest inventory procedure is now to be tested in practice on around 300 sample areas in the Ebensee forest district, both in the productive forest and in the protection forest of the Rindbach catchment area. Furthermore, the automatic evaluation routines are to be further developed. In addition to the traditional inventory parameters, the laser scans will be used to automatically and precisely derive indicators of regeneration density, crown closure for the establishment of fir regeneration, and deadwood stocks.