We enhance photographs shot in dark environments by combining a picture taken with the available light and one flash. preserve ambiance of original lighting insert sharpness from flash image. use bilateral filter to decompose images into detail large scale. reconstruct image using scale detect correct shadows. This combines advantages illumination photography.

Progressive Visual Analytics aims at improving the interactivity in existing analytics techniques by means of visualization as well interaction with intermediate results. One key method for data analysis is dimensionality reduction, example, to produce 2D embeddings that can be visualized and analyzed efficiently. t-Distributed Stochastic Neighbor Embedding (tSNE) a well-suited technique high-dimensional data. tSNE create meaningful results but suffers from slow initialization constrains its...

We propose a new approach to efficiently render large volumetric data sets. The system achieves interactive real-time rendering performance for several billion voxels.

Mass cytometry allows high-resolution dissection of the cellular composition immune system. However, high-dimensionality, large size, and non-linear structure data poses considerable challenges for analysis. In particular, dimensionality reduction-based techniques like t-SNE offer single-cell resolution but are limited in number cells that can be analyzed. Here we introduce Hierarchical Stochastic Neighbor Embedding (HSNE) analysis mass sets. HSNE constructs a hierarchy similarities...

Accurate maps of surface water are essential for many environmental applications. Surface can be generated by combining measurements from multiple sources. Precise estimation using satellite imagery remains a challenging task due to the sensor limitations, complex land cover, topography, and atmospheric conditions. As complementary dataset, in case hilly landscapes, drainage network extracted high-resolution digital elevation models. Additionally, Volunteered Geographic Information (VGI)...

We present a novel rendering system for defocus blur and lens effects. It supports physically-based outperforms previous approaches by involving GPU-based tracing method. Our solution achieves more precision than competing real-time solutions our results are mostly indistinguishable from offline rendering. method is also general can integrate advanced simulations, such as simple geometric models enabling various aberration These latter crucial realism, but often employed in artistic...

Deep neural networks are now rivaling human accuracy in several pattern recognition problems. Compared to traditional classifiers, where features handcrafted, learn increasingly complex directly from the data. Instead of handcrafting features, it is network architecture that manually engineered. The parameters such as number layers or filters per layer and their interconnections essential for good performance. Even though basic design guidelines exist, designing a an iterative...

Abstract Indirect illumination is an important element for realistic image synthesis, but its computation expensive and highly dependent on the complexity of scene BRDF involved surfaces. While off‐line pre‐baking can be acceptable some cases, many applications (games, simulators, etc.) require real‐time or interactive approaches to evaluate indirect illumination. We present a novel algorithm compute lighting in that avoids costly precomputation steps not restricted low‐frequency It based...

Abstract In recent years, dimensionality‐reduction techniques have been developed and are widely used for hypothesis generation in Exploratory Data Analysis. However, these confronted with overcoming the trade‐off between computation time quality of provided dimensionality reduction. this work, we address limitation, by introducing Hierarchical Stochastic Neighbor Embedding (Hierarchical‐SNE). Using a hierarchical representation data, incorporate well‐known mantra Overview‐First,...

Abstract To understand how the immune system works, one needs to have a clear picture of its cellular compositon and cells' corresponding properties functionality. Mass cytometry is novel technique determine single‐cells with unprecedented detail. This amount detail allows for much finer differentiation but also comes at cost more complex analysis. In this work, we present Cytosplore, implementing an interactive workflow analyze mass data in integrated system, providing multiple linked...

This paper presents a novel approach that uses graphics hardware to dynamically calculate voxel-based representation of scene. The voxelization is obtained on run-time in the order milliseconds, even for complex and dynamic scenes containing more than 1,000,000 polygons. created stored GPU avoiding unnecessary data transfer. can handle both regular grids locally optimized better fit scene geometry. demonstrates applications shadow calculation, refraction simultation volume culling/clamping.

Binocular disparity is an important cue for the human visual system to recognize spatial layout, both in reality and simulated virtual worlds. This paper introduces a perceptual model of computer graphics that used define metric compare stereo image alternative estimate magnitude perceived change. Our can be assess effect control level undesirable distortions or enhancements (introduced on purpose). A number psycho-visual experiments are conducted quantify mutual frequency derive model....

We propose a non-permanent add-on that enables plenoptic imaging with standard cameras. Our design is based on physical copying mechanism multiplies sensor image into number of identical copies still carry the information interest. Via different optical filters, we can then recover desired information. A minor modification also allows for aperture sub-sampling and, hence, light-field imaging. As filters in our are exchangeable, reconfiguration purposes possible. show prototype setup high...

Binocular disparity is an important cue for the human visual system to recognize spatial layout, both in reality and simulated virtual worlds. This paper introduces a perceptual model of computer graphics that used define metric compare stereo image alternative estimate magnitude perceived change. Our can be assess effect control level undesirable distortions or enhancements (introduced on purpose). A number psycho-visual experiments are conducted quantify mutual frequency derive model....

We present a system for the annotation and augmentation of mountain photographs. The key issue resides in registration given photograph with 3D geo-referenced terrain model. Typical outdoor images contain little structural information, particularly scenes whose aspect changes drastically across seasons varying weather conditions. Existing approaches usually fail on such difficult scenarios. To avoid burden manual registration, we propose novel automatic technique. Given only viewpoint FOV...

Innate lymphoid cells (ILCs) are abundant in mucosal tissues and involved tissue homeostasis barrier function. Although several ILC subsets have been identified, it is unknown if additional heterogeneity exists, their differentiation pathways remain largely unclear. We applied mass cytometry to analyze ILCs the human fetal intestine distinguished 34 distinct clusters through a t-SNE-based analysis. A lineage (Lin)-CD7+CD127-CD45RO+CD56+ population clustered between CD127+ natural killer (NK)...

We present a method for the real-time simulation of deformable objects that combines robustness, generality, and high performance Projective Dynamics with efficiency scalability offered by model reduction techniques. The decouples cost time integration from mesh resolution can simulate large meshes in real-time. proposed hyper-reduction novel fast approximation constraint projections scalable construction sparse subspace bases. resulting system achieves rates sub-spaces enabling rich...

In recent years the t-distributed Stochastic Neighbor Embedding (t-SNE) algorithm has become one of most used and insightful techniques for exploratory data analysis high-dimensional data. It reveals clusters points at different scales while only requiring minimal tuning its parameters. However, computational complexity limits application to relatively small datasets. To address this problem, several evolutions t-SNE have been developed in years, mainly focusing on scalability similarity...

Developing strategies to mitigate or adapt the threats of floods is an important topic in context climate changes. Many world's cities are endangered due rising ocean levels and changing precipitation patterns. It therefore crucial develop analytical tools that allow us evaluate investigate influence mitigation adaptation measures, such as stronger dikes, adaptive spatial planning, flood disaster plans. Up until present, have only been accessible domain experts, involved simulation processes...

Immersive displays allow presentation of rich video content over a wide field view. We present method to boost visual importance for selected - possibly invisible scene part in cluttered virtual environment. This desirable feature enables unobtrusively guide the gaze direction user any location within immersive 360° surrounding. Our is based on subtle which did not include head rotations previous work. For covering full environment and view, we contribute an approach dynamic stimulus...

This paper is concerned with a fundamental problem in geometric deep learning that arises the construction of convolutional neural networks on surfaces. Due to curvature, transport filter kernels surfaces results rotational ambiguity, which prevents uniform alignment these surface. We propose network architecture for consists vector-valued, rotation-equivariant features. The equivariance property makes it possible locally align features, were computed arbitrary coordinate systems, when...

We enhance photographs shot in dark environments by combining a picture taken with the available light and one flash. preserve ambiance of original lighting insert sharpness from flash image. use bilateral filter to decompose images into detail large scale. reconstruct image using scale detect correct shadows. This combines advantages illumination photography.