A5058354067- Data Management and Algorithms
- Geographic Information Systems Studies
- Biometric Identification and Security
- Human Mobility and Location-Based Analysis
- Data Mining Algorithms and Applications
- Image Retrieval and Classification Techniques
- Data-Driven Disease Surveillance
- Air Traffic Management and Optimization
- Image and Object Detection Techniques
- Mineral Processing and Grinding
- Plant Water Relations and Carbon Dynamics
- Aerospace and Aviation Technology
- Face recognition and analysis
- Traffic Prediction and Management Techniques
- Tree-ring climate responses
- Geology and Paleoclimatology Research
- Anomaly Detection Techniques and Applications
- 3D Modeling in Geospatial Applications
- Video Surveillance and Tracking Methods
- Advanced Image and Video Retrieval Techniques
- Automated Road and Building Extraction
- Urban Transport and Accessibility
- Advanced Aircraft Design and Technologies
- Fire Detection and Safety Systems
- Spatial Cognition and Navigation
Forschungszentrum Jülich
2024-2025
Rosenheim Technical University of Applied Sciences
2024
Darmstadt University of Applied Sciences
2024
University of Tübingen
2024
University of Cologne
2023
Universitas Nusa Bangsa
2023
Fraunhofer Institute for Intelligent Analysis and Information Systems
2007-2021
Google (United States)
2021
Japan Science and Technology Agency
2021
Microsoft (United States)
2021
Quantifying and partitioning evapotranspiration (ET) of agricultural ecosystems in various environmental settings enable studying the site-specific determinants plant water use. The aim study was to conduct reliable reproducible field-scale ET into its component fluxes soil evaporation (E) transpiration (T) from stable isotope analysis (δ2H δ18O). Isotope-based methods, because their methodological independence other, traditional, experimental or data-driven approaches, are...
Triggered by a market relevant application that involves making joint predictions of pedestrian and public transit flows in urban areas, we address the question how to utilize hidden common cause relations among variables interest order improve performance two related regression tasks. Specifically, propose stacked Gaussian process learning, meta-learning scheme which base is enhanced adding posterior covariance functions other tasks its function stage-wise optimization. The idea covariances...
This study aimed to add the limited empirical knowledge base on evidence-based practice (EBP) and social work in Germany. Using a regional sample of 158 workers that were recruited through stratified sampling procedures, examined utilisation as well attitudes toward use research-based methods, administering normed internationally recognised instrument, Evidence-Based Practice Attitude Scale (EBPAS). It further practitioner-level predictors openness methods. Findings indicated familiarity...
The central goal of this project is to establish long-term isotope-specific monitoring H2O, CO2, CH4 and N2O exchange between terrestrial ecosystems the atmosphere with high temporal resolution at grassland, arable land forest sites. H2O CO2 ecosystem will be determined online rapid (10 Hz) laser analyzers using eddy covariance (EC) method. objective measurement determine isotopologue fluxes respective gases source/sink partitioning, i.e., evaporation transpiration in case water vapor or...
Trademarks are recognizable images and/or words used to distinguish various products or services. They become associated with the reputation, innovation, quality, and warranty of products. Countries around world have offices for industrial/intellectual property (IP) registration. A new trademark image in application registration should be distinct from all registered trademarks. Due volume applications size databases containing existing trademarks, it is impossible humans make comparisons...
Web-based geographic information systems (GIS) and planning support are widely adopted as digital tools to practices. The respective solutions tend be isolated implementations aimed at a single purpose due the specific requirement concerning their data, methodology, involved stakeholders, etc. With data platforms, GIS infrastructures, possibility use web-based software that relies on open standards, creating infrastructure is more feasible than ever. Such infrastructures can create...
We introduce s-kNN, a nearest neighbor based spatial data mining algorithm. It belongs to the class of vector-geometry algorithms that reason on complex objects instead point measurements. In contrast most methods in this class, it does fly computations cannot be replaced by pre-processing step without sacrificing efficiency. The key is partial evaluation scheme for efficient computations. algorithm fully integrated into an object-relational database. basis traffic frequency predictions...
This paper describes the results of 2023 edition "LivDet" series iris presentation attack detection (PAD) competitions. New elements in this fifth competition include (1) GAN-generated images as a category instruments (PAI), and (2) an evaluation human accuracy at detecting PAI reference benchmark. Clarkson University Notre Dame contributed image datasets for competition, composed samples representing seven different categories, well baseline PAD algorithms. Fraunhofer IGD, Beijing Civil...
The article discusses the development of a comprehensive modal emissions model that is currently being integrated with variety transportation models as part National Cooperative Highway Research Program project 25-11. Described second-by-second engine-out and tailpipe data collection on 340 light duty vehicles were tested under conditions. variability observed between within groups over different driving modes. initial statistical analysis validation methods are summarized.
In this paper, we discuss an application of spatial data mining to predict pedestrian flow in extensive road networks using a large biased sample. Existing out-of-the-box techniques are not able appropriately deal with its challenges and constraints, particular sample selection bias. For purpose, introduce s-knn-apriori, efficient nearest neighbor based algorithm that allows prior knowledge deductive models be included straightforward easy way.
Abstract The use of small uncrewed aircraft systems (UAS) can effectively capture the wind profile in lower atmospheric boundary layer. This study presents a calibration process to estimate horizontal vector using rotary-wing UAS hovering conditions. procedure does not require tunnels or meteorological masts, only data from flight control unit and specific set flights. A model based on drag coefficient was proposed compared traditional approach. Validation flights at German Weather Service...
Lithological classification is important to improve control of the grinding process in a mining plant. Based on lithological hardness mineral can be estimated and mill operation optimized. In this paper we proposed method for rock based Gabor texture analysis support vector machine classification. We use images from database formed using rocks extracted typical plant Chile. Six different classes were used. Ten each class used training another ten was testing. Results testing measured 5...
Chromatin immunoprecipitation followed by deep sequencing (ChIP-Seq) is a widely used approach to study protein-DNA interactions. Often, the quantities of interest are differential occupancies relative controls, between genetic backgrounds, treatments, or combinations thereof. Current methods for occupancy ChIP-Seq data rely however on binning sliding window techniques, which choice and bin sizes subjective.Here, we present GenoGAM (Genome-wide Generalized Additive Model), brings...
In this work, a No-Reference method for assessing quality of ID Card images is developed, using combination Face Image Quality Assessment and Text Assessment, motivated by the fact that face text are two main components on an Card. For evaluation, new private dataset was created, consisting 12,960 Chilean Cards, with their corresponding printed facial reference image verification. The results obtained proposed assessment method, show as more low discarded, simultaneous improvement in...
Activities developed under hazardous conditions, such as mining, require an active investment from the industry to avoid serious injuries and even loss of lives. One safety measures taken address this problem is implement a video surveillance system able track vehicles workers enforce protocols preventing dangerous events that could lead accidents. In paper, we propose intelligent detect predict possible interactions between operating on monitored environment. Our analyze information one or...
There are many applications in image analysis where it is important to detect accurately patterns that include color and texture, e.g., plastic or concrete traffic barriers. This paper proposes a new method extends general machine vision approach for on-line pattern detection using textural information. Our proposed includes the following steps: division of each into sub-images, use Haralick Binary Quaternion-Moment-Preserving methods extract texture features, support vector machines...