Alexandru Cristian Telea

Utrecht University | UU · Department of Information and Computing Sciences

Physics-based fluid simulation has played an increasingly important role in the computer graphics community. Recent methods in this area have greatly improved the generation of complex visual effects and its computational efficiency. Novel techniques have emerged to deal with complex boundaries, multiphase fluids, gas–liquid interfaces, and fine de...

For vortex particle methods relying on SPH‐based simulations, the direct approach of iterating all fluid particles to capture velocity from vorticity can lead to a significant computational overhead during the Biot‐Savart summation process. To address this challenge, we present a Monte Carlo vortical smoothed particle hydrodynamics (MCVSPH) method...

As the complexity of machine learning (ML) models increases and their application in different (and critical) domains grows, there is a strong demand for more interpretable and trustworthy ML. A direct, model‐agnostic, way to interpret such models is to train surrogate models—such as rule sets and decision trees—that sufficiently approximate the or...

Few-shot colorization aims to learn a model to colorize images with little training data. Yet, existing models often fail to keep color consistency due to ignored patch correlations of the images. In this paper, we propose PCCNet, a novel Patch-wise Contrastive Colorization Network to learn color synthesis by measuring the similarities and variatio...

As the complexity of machine learning (ML) models increases and the applications in different (and critical) domains grow, there is a strong demand for more interpretable and trustworthy ML. One straightforward and model-agnostic way to interpret complex ML models is to train surrogate models, such as rule sets and decision trees, that sufficiently...

Dimensionality reduction (DR) methods create 2D scatterplots of high-dimensional data for visual exploration. As such scatterplots are often used to reason about the cluster structure of the data, this requires DR methods with good cluster preservation abilities. Recently, Sharpened DR (SDR) was proposed to enhance the ability of existing DR method...

We present a conceptual framework for the development of visual interactive techniques to formalize and externalize trust in machine learning (ML) workflows. Currently, trust in ML applications is an implicit process that takes place in the user’s mind. As such, there is no method of feedback or communication of trust that can be acted upon. Our fr...

Multivariate functions have a central place in the development of techniques present many domains, such as machine learning and optimization research. However, only a few visual techniques exist to help users understand such multivariate problems, especially in the case of functions that depend on complex algorithms and variable constraints. In thi...

Understanding how a machine learning classifier works is an important task in machine learning engineering. However, doing this is for any classifier in general difficult. We propose to leverage visualization methods for this task. For this, we extend a recent technique called Decision Boundary Map (DBM) which graphically depicts how a classifier p...

Dimensionality reduction (DR) is an essential tool for the visualization of high-dimensional data. The recently proposed Self-Supervised Network Projection (SSNP) method addresses DR with a number of attractive features, such as high computational scalability, genericity, stability and out-of-sample support, computation of an inverse mapping, and t...

Applying dimensionality reduction (DR) to large, high-dimensional data sets can be challenging when distinguishing the underlying high-dimensional data clusters in a 2D projection for exploratory analysis. We address this problem by first sharpening the clusters in the original high-dimensional data prior to the DR step using Local Gradient Cluster...

Projection techniques are often used to visualize high-dimensional data, allowing users to better understand the overall structure of multi-dimensional spaces on a 2D screen. Although many such methods exist, comparably little work has been done on generalizable methods of inverse-projection -- the process of mapping the projected points, or more g...

Applying dimensionality reduction (DR) to large, high-dimensional data sets can be challenging when distinguishing the underlying high-dimensional data clusters in a 2D projection for exploratory analysis. We address this problem by first sharpening the clusters in the original high-dimensional data prior to the DR step using Local Gradient Cluster...

Training deep neural networks is challenging when large and annotated datasets are unavailable. Extensive manual annotation of data samples is time-consuming, expensive, and error-prone, notably when it needs to be done by experts. To address this issue, increased attention has been devoted to techniques that propagate uncertain labels (also called...

Projections aim to convey the relationships and similarity of high‐dimensional data in a low‐dimensional representation. Most such techniques are designed for static data. When used for time‐dependent data, they usually fail to create a stable and suitable low dimensional representation. We propose two dynamic projection methods (PCD‐tSNE and LD‐tS...

Automatic detection of brain anomalies in MR images is challenging and complex due to intensity similarity between lesions and healthy tissues as well as the large variability in shape, size, and location among different anomalies. Even though discriminative models (supervised learning) are commonly used for this task, they require quite high-quali...

Expert human supervision of the large labeled training sets needed by convolutional neural networks is expensive. To obtain sufficient labeled samples to train a model, one can propagate labels from a small set of supervised samples to a large unsupervised set. Yet, such methods need many supervised samples for validation. We present a method that...

Data annotation using visual inspection (supervision) of each training sample can be laborious. Interactive solutions alleviate this by helping experts propagate labels from a few supervised samples to unlabeled ones based solely on the visual analysis of their feature space projection (with no further sample supervision). We present a semi-automat...

A major issue in smoothed particle hydrodynamics (SPH) approaches is the numerical dissipation during the projection process, especially under coarse discretizations. High‐frequency details, such as turbulence and vortices, are smoothed out, leading to unrealistic results. To address this issue, we introduce a vorticity refinement (VR) solver for S...

A major issue in Smoothed Particle Hydrodynamics (SPH) approaches is the numerical dissipation during the projection process, especially under coarse discretizations. High-frequency details, such as turbulence and vortices, are smoothed out, leading to unrealistic results. To address this issue, we introduce a Vorticity Refinement (VR) solver for S...

While convolutional neural networks need large labeled sets for training images, expert human supervision of such datasets can be very laborious. Proposed solutions propagate labels from a small set of supervised images to a large set of unsupervised ones to obtain sufficient truly-and-artificially labeled samples to train a deep neural network mod...

Several brain disorders are associated with abnormal brain asymmetries (asymmetric anomalies). Several computer-based methods aim to detect such anomalies automatically. Recent advances in this area use automatic unsupervised techniques that extract pairs of symmetric supervoxels in the hemispheres, model normal brain asymmetries for each pair from...

Data annotation using visual inspection (supervision) of each training sample can be laborious. Interactive solutions alleviate this by helping experts propagate labels from a few supervised samples to unlabeled ones based solely on the visual analysis of their feature space projection (with no further sample supervision). We present a semi-automat...

Data annotation using visual inspection (supervision) of each training sample can be laborious. Interactive solutions alleviate this by helping experts propagate labels from a few supervised samples to unlabeled ones based solely on the visual analysis of their feature space projection (with no further sample supervision). We present a semi-automat...

The reconstruction of phase spaces is an essential step to analyze time series according to Dynamical System concepts. A regression performed on such spaces unveils the relationships among system states from which we can derive their generating rules, that is, the most probable set of functions responsible for generating observations along time. In...

Dimensionality reduction methods are an essential tool for multidimensional data analysis, and many interesting processes can be studied as time‐dependent multivariate datasets. There are, however, few studies and proposals that leverage on the concise power of expression of projections in the context of dynamic/temporal data. In this paper, we aim...

Rectangular treemaps are often the method of choice to visualize large hierarchical datasets. Nowadays such datasets are available over time, hence there is a need for (a) treemaps that can handle time‐dependent data, and (b) corresponding quality criteria that cover both a treemap's visual quality and its stability over time. In recent years a wid...

Animated visualizations are one of the methods for finding and understanding complex structures of time‐dependent vector fields. Many visualization designs can be used to this end, such as streamlines, vector glyphs, and image‐based techniques. While all such designs can depict any vector field, their effectiveness in highlighting particular field...

Dimensionality reduction methods, also known as projections, are often used to explore multidimensional data in machine learning, data science, and information visualization. However, several such methods, such as the well-known t-distributed stochastic neighbor embedding and its variants, are computationally expensive for large datasets, suffer fr...

Skeletons are well-known descriptors that capture the geometry and topology of 2D and 3D shapes. We leverage these properties by using surface skeletons to remove noise from 3D shapes. For this, we extend an existing method that removes noise, but keeps important (salient) corners for 2D shapes. Our method detects and removes large-scale, complex,...

Dimensionality reduction methods are an essential tool for multidimensional data analysis, and many interesting processes can be studied as time-dependent multivariate datasets. There are, however, few studies and proposals that leverage on the concise power of expression of projections in the context of dynamic/temporal data. In this paper, we aim...

Multidimensional projections (MPs) are established tools for exploring the structure of high-dimensional datasets to reveal groups of similar observations. For optimal usage, MPs can be augmented with mechanisms that explain what such points have in common that makes them similar. We extend the set of such explanatory instruments by two new techniq...

Dimensionality reduction methods, also known as projections, are frequently used in multidimensional data exploration in machine learning, data science, and information visualization. Tens of such techniques have been proposed, aiming to address a wide set of requirements, such as ability to show the high-dimensional data structure, distance or nei...

Visualizing decision boundaries of machine learning classifiers can help in classifier design, testing and fine-tuning. Decision maps are visualization techniques that overcome the key sparsity-related limitation of scatterplots for this task. To increase the trustworthiness of decision map use, we perform an extensive evaluation considering the di...

New solutions in operational environments are often, among objective measurements, evaluated by using subjective assessment and judgment from experts. Anyhow, it has been demonstrated that subjective measures suffer from poor resolution due to a high intra and inter-operator variability. Also, performance measures, if available, could provide just...

Most neurological diseases are associated with abnormal brain asymmetries. Recent advances in automatic unsupervised techniques model normal brain asymmetries from healthy subjects only and treat anomalies as outliers. Outlier detection is usually done in a common standard coordinate space that limits its usability. To alleviate the problem, we ext...

Rectangular treemaps are often the method of choice for visualizing large hierarchical datasets. Nowadays such datasets are available over time, and hence there is a need for (a) treemapping algorithms that can handle time-dependent data, and (b) corresponding quality criteria and evaluations. In recent years a wide variety of treemapping algorithm...

Origin‐destination (OD) trails describe movements across space. Typical visualizations thereof use either straight lines or plot the actual trajectories. To reduce clutter inherent to visualizing large OD datasets, bundling methods can be used. Yet, bundling OD trails in urban traffic data remains challenging. Two specific reasons hereof are the co...

RadViz is one of the few methods in Visual Analytics able to project high-dimensional data and explain formed structures in terms of data variables. However, RadViz methods have several limitations in terms of scalability in the number of variables, ambiguities created in the projection by the placement of variables along the circular design space,...

Dimensionality reduction methods, also known as projections, are frequently used for exploring multidimensional data in machine learning, data science, and information visualization. Among these, t-SNE and its variants have become very popular for their ability to visually separate distinct data clusters. However, such methods are computationally e...

Although deep neural networks have achieved state-of-the-art performance in several artificial intelligence applications in the past decade, they are still hard to understand. In particular, the features learned by deep networks when determining whether a given input belongs to a specific class are only implicitly described concerning a considerabl...

While the number of unsupervised samples for data annotation is usually high, the absence of large supervised train- ing sets for effective feature learning and design of high-quality classifiers is a known problem whenever specialists are required for data supervision. By exploring the feature space of supervised and unsupervised samples, semi-sup...

This study aims at investigating the possibility to employ neurophysiological measures to assess the humanmachine interaction effectiveness. Such a measure can be used to compare new technologies or solutions, with the final purpose to enhance operator's experience and increase safety. In the present work, two different interaction modalities (Norm...

Automation of Software Development Life Cycle (SDLC) requires continuous verification of compliance of the software product under construction to a set of expectations about its quality. We define a policy as an expectation about some aspect of software quality that is expressed in a form of an array of non-functional requirements (NFR), compliance...

Understanding large multidimensional datasets is one of the most challenging problems in visual data exploration. One key challenge that increases the size of the exploration space is the number of views that one can generate from a single dataset, based on the use of multiple parameter values and exploration paths. Often, no such single view conta...

Bundling techniques provide a visual simplification of a graph drawing or trail set, by spatially grouping similar graph edges or trails. This way, the structure of the visualization becomes simpler and thereby easier to comprehend in terms of assessing relations that are encoded by such paths, such as finding groups of strongly interrelated nodes...

We present a set of visualization methods for the analysis of multivariate data recorded from the measurement of the performance of athletes during training. We use a modified training device to measure the force, acceleration, displacement, and speed of the athlete’s feet and arms while performing a certain training exercise. We are interested in...

Tool support for program understanding becomes increasingly important in the software evolution cycle, and it has become an integral part of managing systems evolution and maintenance. Using interactive visual tools for getting insight into large evolving legacy information systems has gained popularity. Although several such tools exist, few of th...