A5090930477- Advanced Image and Video Retrieval Techniques
- Computer Graphics and Visualization Techniques
- Image Retrieval and Classification Techniques
- Advanced Vision and Imaging
- Virtual Reality Applications and Impacts
- Data Visualization and Analytics
- Robotics and Sensor-Based Localization
- 3D Shape Modeling and Analysis
- Human Pose and Action Recognition
- Generative Adversarial Networks and Image Synthesis
- Augmented Reality Applications
- Image Processing and 3D Reconstruction
- Hand Gesture Recognition Systems
- Multimodal Machine Learning Applications
- Manufacturing Process and Optimization
- Acute Ischemic Stroke Management
- AI in cancer detection
- Advanced Neural Network Applications
- Advanced Surface Polishing Techniques
- Additive Manufacturing and 3D Printing Technologies
- Advanced Image Processing Techniques
- Anxiety, Depression, Psychometrics, Treatment, Cognitive Processes
- Nanofabrication and Lithography Techniques
- Domain Adaptation and Few-Shot Learning
- Radiomics and Machine Learning in Medical Imaging
Berliner Hochschule für Technik
2017-2024
Technische Universität Berlin
2009-2017
Hillenbrand (United States)
2017
Centre Inria de l'Université de Lorraine
2017
Institut national de recherche en informatique et en automatique
2017
Laboratoire Lorrain de Recherche en Informatique et ses Applications
2017
University of British Columbia
2006
Bauhaus-Universität Weimar
2003-2006
Max Planck Society
2006
We introduce a benchmark for evaluating the performance of large-scale sketch-based image retrieval systems. The necessary data are acquired in controlled user study where subjects rate how well given sketch/image pairs match. suggest to use as large database made publicly available further studies this type. Furthermore, we develop new descriptors based on bag-of-features approach and demonstrate that they significantly outperform other literature.
We develop a system for 3D object retrieval based on sketched feature lines as input. For objective evaluation, we collect large number of query sketches from human users that are related to an existing data base objects. The turn out be generally quite abstract with local and global deviations the original shape. Based this observation, decide use bag-of-features approach over computer generated line drawings targeted transform Gabor filters system. can show objectively is better suited...
Brain vessel status is a promising biomarker for better prevention and treatment in cerebrovascular disease. However, classic rule-based segmentation algorithms need to be hand-crafted are insufficiently validated. A specialized deep learning method-the U-net-is alternative. Using labeled data from 66 patients with disease, the U-net framework was optimized evaluated three metrics: Dice coefficient, 95% Hausdorff distance (95HD) average (AVD). The model performance compared traditional...
We address the problem of large scale sketch based image retrieval, searching in a database over million images. The search is on descriptor that elegantly addresses asymmetry between binary user one hand and full color other hand. proposed constructed such both undergo exactly same preprocessing steps. also design an adapted version for MPEG-7 compare their performance 1.5 Best matching images are clustered histograms, to offset lacking query. Overall, query results demonstrate system...
Photosketcher is an interactive system for progressively synthesizing novel images using only sparse user sketches as input. works on the image content exclusively; it doesn't require keywords or other metadata associated with images. Users sketch rough shape of a desired part, and searches large collection it. The search based bag-of-features approach that uses local descriptors translation-invariant retrieval parts. Composition scribbles: from scribbles, predicts part Gaussian mixture...
Abstract We introduce an algorithm and representation for fabricating 3D shape abstractions using mutually intersecting planar cut‐outs. The planes have prefabricated slits at their intersections are assembled by sliding them together. Often such used as a sculptural art form or in architecture colloquially called ‘cardboard sculptures’. Based on analysis of construction rules, we propose extended binary space partitioning tree efficient cardboard models which allows us to quickly evaluate...
We develop a system for 3D object retrieval based on sketched feature lines as input. For objective evaluation, we collect large number of query sketches from human users that are related to an existing data base objects. The turn out be generally quite abstract with local and global deviations the original shape. Based this observation, decide use bag-of-features approach over computer generated line drawings targeted transform Gabor filters system. can show objectively is better suited...
Deep learning requires large labeled datasets that are difficult to gather in medical imaging due data privacy issues and time-consuming manual labeling. Generative Adversarial Networks (GANs) can alleviate these challenges enabling synthesis of shareable data. While 2D GANs have been used generate images with their corresponding labels, they cannot capture the volumetric information 3D imaging. more suitable for this volumes but not labels. One reason might be synthesizing is challenging...
Sketch-based 3D shape retrieval has become an important research topic in content-based object retrieval. The aim of this track is to measure and compare the performance sketch-based methods implemented by different participants over world. based on a new benchmark, which contains two types sketch queries versions target models. In track, 7 runs have been submitted 5 groups their accuracies were evaluated using commonly used metrics. We hope that its corresponding evaluation code,...
We introduce a system for progressively creating images through simple sketching and compositing interface. A large database of over 1.5 million is searched matches to user's binary outline sketch; the results this search can be combined interactively synthesize desired image. image descriptors task estimating difference between sketches. The part based on graph cut Poisson blending. demonstrate that resulting allows generating complex in an intuitive way.
Slicing is the procedure necessary to prepare a shape for layered manufacturing. There are degrees of freedom in this process, such as starting point slicing sequence and thickness each slice. The choice these parameters influences manufacturing process its result: number slices significantly affects time needed manufacturing, while their error. Assuming discrete setting, we measure error voxels that incorrectly assigned due slicing. We provide an algorithm generates, given set available...
As large collections of 3D models are starting to become as common public image collections, the need arises quickly locate in such collections. Models often insufficiently annotated that a keyword based search is not promising. Our approach for content searching relies entirely on visual analysis and observation part our perception shapes stems from their salient features, usually captured by dominant lines display. Recent research feature has shown 1) people mostly draw same when asked...
Slicing is the procedure necessary to prepare a shape for layered manufacturing. There are degrees of freedom in this process, such as starting point slicing sequence and thickness each slice. The choice these parameters influences manufacturing process its result: number slices significantly affects time needed manufacturing, while their error. Assuming discrete setting, we measure error voxels that incorrectly assigned due slicing. We provide an algorithm generates, given set available...
Stars form in dense clouds of interstellar gas and dust. The residual dust surrounding a young star scatters diffuses its light, making the star's "cocoon" observable from Earth. resulting structures, called reflection nebulae, are commonly very colorful appearance due to wavelength-dependent effects scattering extinction light. intricate interplay cause color hues, brightness distributions, apparent shapes such nebulae vary greatly with viewpoint. We describe an interactive visualization...
Virtual reality (VR) is becoming increasingly integrated into people’s lives. However, current applications are usually limited to a certain fixed location. With the introduction of wireless and mobile head mounted displays (HMD) there no longer need for workstation. In this way it possible introduce VR experiences in public spaces. Some airlines have already started use as an alternative entertainment system on flights.
Acquiring panoramic images using stitching takes a lot of time and moving objects may cause ghosting. It is also difficult to obtain full spherical panorama, because the downward picture cannot be captured while camera mounted on tripod.
With a significant improvement in virtual reality (VR) devices, the number of interaction-based applications for consumers and industrial products is naturally increasing. As result, many people can use VR their homes or offices where they are limited by physical tracking space. One way to overcome this limitation natural walking perception-inspired locomotion techniques such as redirected (RDW). RDW utilizes imperfections human perception introduce small changes like rotations translations...
Sharing labeled data is crucial to acquire large datasets for various Deep Learning applications. In medical imaging, this often not feasible due privacy regulations. Whereas anonymization would be a solution, standard techniques have been shown partially reversible. Here, synthetic using Generative Adversarial Network (GAN) with differential guarantees could solution ensure the patient's while maintaining predictive properties of data. study, we implemented Wasserstein GAN (WGAN) and...
Virtual and augmented reality (VR/AR) exergames are promising tools for increasing training motivation. However, the use of with mixed (MR) headsets remains under-researched. Older adults hypertension could also benefit from increased adherence associated MR. Endurance strength endurance exercises recommended this group to lower blood pressure. The aim preliminary study (n = 22) was compare usability acceptance two exergames, which represent different types—strength (SET) (ET). developed...
Stroke is a major cause of death or disability. As imaging-based patient stratification improves acute stroke therapy, dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI) interest in image brain perfusion. However, expert-level perfusion maps require manual semi-manual post-processing by medical expert making the procedure time-consuming and less-standardized. Modern machine learning methods such as generative adversarial networks (GANs) have potential to automate map...
Background Virtual reality exposure therapy (VRET) is a promising treatment approach for anxiety disorders. However, while its efficacy has been demonstrated in adults, research on the of VRET adolescents with disorders largely lacking. Objective A pilot study was carried out to test whether virtual (VR) school environment elicits state and autonomic arousal (diagnoses covering social disorder or specific phobia involving contexts). In addition, we examined repeated VR led reduction this...