Michal Kawulok

Michal Kawulok
Silesian University of Technology · Institute of Computer Science

Associate Professor

About

150
Publications
19,564
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
1,773
Citations
Additional affiliations
April 2018 - present
Silesian University of Technology
Position
  • Professor (Associate)
June 2012 - present
Future Processing
Position
  • Senior Researcher
May 2007 - March 2018
Silesian University of Technology
Position
  • Professor (Assistant)

Publications

Publications (150)
Article
Full-text available
In this paper we propose a new method for skin detection in color images which consists in spatial analysis using the introduced texture-based discriminative skin-presence features. Color-based skin detection has been widely explored and many skin color modeling techniques were developed so far. However, efficacy of the pixel-wise classification is...
Article
Full-text available
In this paper we present how to exploit the textural information to improve scribble-based image colorization. Although many methods have been already proposed for coloring grayscale images based on a set of color scribbles inserted by a user, very few of them take into account textural properties. We demonstrate that the textural information can b...
Chapter
This chapter presents an overview of existing methods for human skin detection and segmentation. First of all, the skin color modeling schemes are outlined, and their limitations are discussed based on the presented experimental study. Then, we explain the techniques which were reported helpful in improving the efficacy of color-based classificatio...
Conference Paper
Full-text available
This paper presents a new method for selecting valuable training data for support vector machines (SVM) from large, noisy sets using a genetic algorithm (GA). SVM training data selection is a known, however not extensively investigated problem. The existing methods rely mainly on analyzing the geometric properties of the data or adapt a randomized...
Article
This paper shows how to improve holistic face analysis by assigning importance factors to different facial regions (termed as face relevance maps). We propose a novel supervised learning algorithm for generating face relevance maps to improve the discriminating capability of existing methods. We have successfully applied the developed technique to...
Preprint
Maintaining farm sustainability through optimizing the agricultural management practices helps build more planet-friendly environment. The emerging satellite missions can acquire multi- and hyperspectral imagery which captures more detailed spectral information concerning the scanned area, hence allows us to benefit from subtle spectral features du...
Preprint
Hyperspectral unmixing remains one of the most challenging tasks in the analysis of such data. Deep learning has been blooming in the field and proved to outperform other classic unmixing techniques, and can be effectively deployed onboard Earth observation satellites equipped with hyperspectral imagers. In this letter, we follow this research path...
Article
Full-text available
Hyperspectral images capture very detailed information about scanned objects and, hence, can be used to uncover various characteristics of the materials present in the analyzed scene. However, such image data are difficult to transfer due to their large volume, and generating new ground-truth datasets that could be utilized to train supervised lear...
Article
Active collision avoidance has become an important task in space operations nowadays, and hundreds of alerts corresponding to close encounters of a satellite and other space objects are typically issued for a satellite in Low Earth Orbit every week. Such alerts are provided in the form of conjunction data messages, and only about two actionable ale...
Article
Support vector machines (SVMs) have been exploited in a plethora of real-life classification and regression tasks, and are one of the most researched supervised learners. However, their generalization abilities strongly depend on the pivotal hyperparameters of the classifier, alongside its training dataset. Also, the training process is computation...
Article
Full-text available
Although hyperspectral images capture very detailed information about the scanned objects, their efficient analysis, transfer, and storage are still important practical challenges due to their large volume. Classifying and segmenting such imagery are the pivotal steps in virtually all applications, hence developing new techniques for these tasks is...
Chapter
Capturing, storing, and analyzing high-dimensional time series data are important challenges that need to be effectively tackled nowadays, as the extremely large amounts of such data are being generated every second. In this paper, we introduce the recurrent neural networks equipped with attention modules that quantify the importance of features, h...
Chapter
Capturing, transferring, and storing high-resolution images has become a serious issue in a wide range of fields, in which these processes are costly, time-consuming, or even infeasible. As obtaining low-resolution images may be easier in practice, enhancing their spatial resolution is currently an active research area and encompasses both single-...
Chapter
Automated brain tumor segmentation is a vital topic due to its clinical applications. We propose to exploit a lightweight U-Net-based deep architecture called Skinny for this task—it was originally employed for skin detection from color images, and benefits from a wider spatial context. We train multiple Skinny networks over all image planes (axial...
Article
Full-text available
Applying computer vision techniques to distinguish between spontaneous and posed smiles is an active research topic of affective computing. Although there have been many works published addressing this problem and a couple of excellent benchmark databases created, the existing state-of-the-art approaches do not exploit the action units defined with...
Article
Although deep learning is gaining more widespread use in hyperspectral image analysis, it is challenging to train high-capacity models in a supervised way--ground-truth sets are expensive to obtain, and they are practically always extremely imbalanced. To deal with the problem of missing ground-truth data, its high dimensionality and potential redu...
Chapter
Support vector machine (SVM) is a well-known machine learning algorithm widely used for classification and regression problems. Despite the high prediction rate of this technique in a wide range of real applications, the efficiency of SVM and its classification performance highly depends on the hyperparameters setting as well as the selection of fe...
Article
Full-text available
Data augmentation is a popular technique which helps improve generalization capabilities of deep neural networks, and can be perceived as implicit regularization. It plays a pivotal role in scenarios in which the amount of high-quality ground-truth data is limited, and acquiring new examples is costly and time-consuming. This is a very common probl...
Conference Paper
Hyperspectral satellite imaging has been gaining enormous research attention due to the latest advancements in the sensor technology, and the amount of information it conveys. However, its efficient analysis, transfer, and storage are still big practical issues which need to be endured in on-board applications. In this paper, we verify if the simul...
Article
Full-text available
Background: Nowadays, not only are single genomes commonly analyzed, but also metagenomes, which are sets of, DNA fragments (reads) derived from microbes living in a given environment. Metagenome analysis is aimed at extracting crucial information on the organisms that have left their traces in an investigated environmental sample.In this study we...
Article
Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) plays an important role in diagnosis and grading of brain tumors. Although manual DCE biomarker extraction algorithms boost the diagnostic yield of DCE-MRI by providing quantitative information on tumor prognosis and prediction, they are time-consuming and prone to human errors. In this...
Article
Super-resolution (SR) reconstruction is a process aimed at enhancing the spatial resolution of images, either from a single observation, based on the learned relation between low and high resolution, or from multiple images presenting the same scene. SR is particularly important, if it is not feasible to acquire images at the desired resolution, wh...
Article
Deep learning has established the state of the art in multiple fields, including hyperspectral image analysis. However, training large-capacity learners to segment such imagery requires representative training sets. Acquiring such data is human-dependent and time-consuming, especially in earth observation scenarios, where the hyperspectral data tra...
Article
Full-text available
Support vector machines (SVMs) are a supervised classifier successfully applied in a plethora of real-life applications. However, they suffer from the important shortcomings of their high time and memory training complexities, which depend on the training set size. This issue is especially challenging nowadays, since the amount of data generated ev...
Preprint
Hyperspectral imaging provides detailed information about the scanned objects, as it captures their spectral characteristics within a large number of wavelength bands. Classification of such data has become an active research topic due to its wide applicability in a variety of fields. Deep learning has established the state of the art in the area,...
Preprint
Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) plays an important role in diagnosis and grading of brain tumor. Although manual DCE biomarker extraction algorithms boost the diagnostic yield of DCE-MRI by providing quantitative information on tumor prognosis and prediction, they are time-consuming and prone to human error. In this p...
Conference Paper
Support vector machine (SVM) classifiers can cope with many different classification tasks but improperly selected hyperparameters may deteriorate their performance. Moreover, datasets are getting bigger in terms of their size and the number of features. This is often coupled with low training data quality and presence of redundant features, which...
Article
Data augmentation helps improve generalization capabilities of deep neural networks when only limited ground-truth training data are available. In this letter, we propose test-time augmentation of hyperspectral data, which is executed during the inference rather than before the training of deep networks. We introduce two augmentation techniques, wh...
Preprint
Deep learning has established the state of the art in multiple fields, including hyperspectral image analysis. However, training large-capacity learners to segment such imagery requires representative training sets. Acquiring such data is human-dependent and time-consuming, especially in Earth observation scenarios, where the hyperspectral data tra...
Preprint
Full-text available
The capabilities of super-resolution reconstruction (SRR)---techniques for enhancing image spatial resolution---have been recently improved significantly by the use of deep convolutional neural networks. Commonly, such networks are learned using huge training sets composed of original images alongside their low-resolution counterparts, obtained wit...
Article
Background and objective: Magnetic resonance imaging (MRI) is an indispensable tool in diagnosing brain-tumor patients. Automated tumor segmentation is being widely researched to accelerate the MRI analysis and allow clinicians to precisely plan treatment-accurate delineation of brain tumors is a critical step in assessing their volume, shape, bou...
Chapter
Super-resolution reconstruction (SRR) is aimed at increasing image spatial resolution from multiple images presenting the same scene or from a single image based on the learned relation between low and high resolution. Emergence of deep learning allowed for improving single-image SRR significantly in the last few years, and a variety of deep convol...
Chapter
Creating classification ensembles may be perceived as a regularization technique which aims at improving the generalization capabilities of a classifier. In this paper, we introduce a multi-level memetic algorithm for evolving classification ensembles (they can be either homo- or heterogeneous). First, we evolve the content of such ensembles, and t...
Preprint
Data augmentation is a popular technique which helps improve generalization capabilities of deep neural networks. It plays a pivotal role in remote-sensing scenarios in which the amount of high-quality ground truth data is limited, and acquiring new examples is costly or impossible. This is a common problem in hyperspectral imaging, where manual an...
Preprint
Super-resolution reconstruction (SRR) is a process aimed at enhancing spatial resolution of images, either from a single observation, based on the learned relation between low and high resolution, or from multiple images presenting the same scene. SRR is particularly valuable, if it is infeasible to acquire images at desired resolution, but many im...
Article
Hyperspectral satellite imaging attracts enormous research attention in the remote sensing community, and hence, automated approaches for precise segmentation of such imagery are being rapidly developed. In this letter, we share our observations on the strategy for validating hyperspectral image segmentation algorithms currently followed in the lit...
Chapter
Support vector machine (SVM) is a popular classifier that has been used to solve a broad range of problems. Unfortunately, its applications are limited by computational complexity of training which is \(O(t^3)\), where t is the number of vectors in the training set. This limitation makes it difficult to find a proper model, especially for non-linea...
Preprint
Hyperspectral satellite imaging attracts enormous research attention in the remote sensing community, hence automated approaches for precise segmentation of such imagery are being rapidly developed. In this letter, we share our observations on the strategy for validating hyperspectral image segmentation algorithms currently followed in the literatu...
Conference Paper
Support vector machines (SVMs) are a well-established classifier, already applied in a variety of pattern recognition tasks. However, they suffer from several drawbacks---selecting their appropriate hyper-parameter values (the SVM model) along with the training sets being the most important. In this paper, we study the influence of applying various...
Article
Super-resolution reconstruction (SRR) consists in enhancing image spatial resolution given a single image or a bunch of images presenting the same scene at lower resolution. Potential benefits of applying SRR to satellite imagery are evident, as it may enhance the capacities of images characterized with lower resolution, shorter revisit times, and...
Conference Paper
Super-resolution reconstruction (SRR) allows for enhancing image spatial resolution from low-resolution (LR) observations, which are assumed to have been derived from a hypothetical high-resolution image by applying a certain imaging model (IM). However, if the actual degradation is different from the assumed IM, which is often the case in real-wor...
Conference Paper
Deep learning is a widely explored research area, as it established the state of the art in many fields. However, the effectiveness of deep neural networks (DNNs) is affected by several factors related with their training. The commonly used gradient-based back-propagation algorithm suffers from a number of shortcomings, such as slow convergence, di...
Poster
Full-text available
Feature extraction is the first step in building real-life classification engines---it aims at elaborating features to characterize objects that are to be labeled by a trained model. Time-consuming feature extraction requires domain expertise to effectively design features. Deep neural networks (DNNs) appeared as a remedy in this context---their sh...
Chapter
Feature extraction is the first step in building real-life classification engines—it aims at elaborating features to characterize objects that are to be labeled by a trained model. Time-consuming feature extraction requires domain expertise to effectively design features. Deep neural networks (DNNs) appeared as a remedy in this context—their shallo...
Chapter
Super-resolution reconstruction (SRR) allows for producing a high-resolution (HR) image from a set of low-resolution (LR) observations. The majority of existing methods require tuning a number of hyper-parameters which control the reconstruction process and configure the imaging model that is supposed to reflect the relation between high and low re...
Poster
Full-text available
Imaging technologies have developed rapidly over the past decade proving to be valuable and effective tools for diagnosis, evaluation and treatment of many conditions, especially cancer. Dynamic contrast enhanced imaging using computed tomography or magnetic resonance has been shown particularly effective and has been intensively studied to allow f...
Conference Paper
Automatic detection of lung lesions from computed tomography (CT) and positron emission tomography (PET) is an important task in lung cancer diagnosis. While CT scans make it possible to retrieve structural information, PET images reveal the functional aspects of the tissue, hence combined PET/CT imagery allows for detecting metabolically active le...
Conference Paper
Super-resolution reconstruction (SRR) consists in enhancing image spatial resolution given a single image or a bunch of images presenting the same scene. Potential benefits of SRR are evident, when images of high resolution are required, but are unavailable due to technological limitations or economic reasons. Obviously, this is inherent to satelli...
Poster
Full-text available
Automatic detection of lung lesions from computed tomography (CT) and positron emission tomography (PET) is an important task in lung cancer diagnosis. While CT scans make it possible to retrieve structural information, PET images reveal the functional aspects of the tissue, hence combined PET/CT imagery allows for detecting metabolically active le...
Conference Paper
Deep neural networks (DNNs) have achieved unprecedented success in a wide array of tasks. However, the performance of these systems depends directly on their hyper-parameters which often must be selected by an expert. Optimizing the hyper-parameters remains a substantial obstacle in designing DNNs in practice. In this work, we propose to select the...
Patent
Method for multimodal visual analysis for measurements of the visual attention of the consumers of multimedia contents in the collective transport vehicles and the system for counting the consumers of multimedia contents in the collective transport vehicles (for details, see http://regserv.uprp.pl/register/application?lng=de&number=P.415337).
Conference Paper
Support vector machines (SVMs) have been found highly helpful in solving numerous pattern recognition tasks. Although it is challenging to train SVMs from large data sets, this obstacle may be mitigated by selecting a small, yet representative, subset of the entire training set. Another crucial and deeply-investigated problem consists in selecting...
Conference Paper
Image segmentation is an initial, yet crucial procedure in a number of medical imaging systems. Despite the existence of numerous generic solutions that address this problem, there is still a need for developing fast and accurate techniques specialized at extracting particular organs from the CT scans. In this paper, we present an approach based on...