Computer Science Journal (AGH University of Science and Technology, Krakow)
Not a member yet
476 research outputs found
Sort by
A new genetic algorithm based on dissimilarities and similarities
No full textOptimization is essential for nding suitable answers to real life problems. In particular, genetic (or more generally, evolutionary) algorithms can provide satisfactory approximate solutions to many problems to which exact analytcal results are not accessible. In this paper we present both theoretical and experimental results on a new genetic algorithm called Dissimilarity and Simlarity of Chromosomes (DSC). This methodology constructs new chromosomes starting with the pairs of existing ones by exploring their dissimilarities and similarities. To demonstrate the performance of the algorithm, it is run on 17 two-dimensional, one four-dimensional and two ten-dimensional optimization problems described in the literature, and compared with the well-known GA, CMA-ES and DE algorithms.The results of tests show the superiority of our strategy in the majority of cases
Lifelogging system based on Averaged Hidden Markov Models: dangerous activities recognition for caregivers support
Get PDFIn this paper a prototype lifelogging system for monitoring persons with cognitive disabilities and elderly people, as well as a method for automatic detection of dangerous activities are presented. The system allows remote monitoring of observed persons via Internet website and respects the privacy of the persons by displaying their silhouettes instead of actual images. Application allows viewing of both real-time and historic data. Lifelogging data (skeleton coordinates) needed for posture and activity recognition are acquired using Microsoft Kinect 2.0. Several activities are marked as potentially dangerous and generate alarms sent to the caregivers upon detection. Recognition models are developed using Averaged Hidden Markov Models with multiple learning sequences. Action recognition includes methods for differentiation between normal and potentially dangerous activities e.g. self-aggressive autistic behavior) using the same motion trajectory. Some activity recognition examples and results are presented
COMPACT: biometric dataset of face images acquired in uncontrolled indoor environment
Get PDFBiometric databases are important components that help to improve state-of-the-art recognition performance. The availability of more and more difficult data attracts the researchers\u27 attention, who systematically develop novel recognition algorithms and increase identification accuracy. Surprisingly, most of the popular face datasets, like LFW or IJBA are not fully unconstrained. The majority of the available images were not acquired on-the-move, which reduces the amount of blur caused by motion or incorrect focusing. Therefore, in this paper, the COMPACT database for studying less-cooperative face recognition is introduced. The dataset consists of high-resolution images of 108 subjects acquired in a fully automated manner as people go through the recognition gate. This ensures that the collected data contains the real world degradation factors: different distances, expressions, occlusions, pose variations and motion blur. Additionally, the authors conducted a series of experiments that verify face recognition performance on the collected data
Hand posture recognition using modified Ensemble of Shape Functions and Global Radius-based Surface Descriptor
Get PDFThis paper presents an approach to recognition of static hand gestures based on data acquired from 3D cameras and point cloud descriptors: Ensemble of Shape Functions and Global Radius-based Surface Descriptor. We describe the recognition algorithm consisting of: hand segmentation, noise removal and downsampling of point cloud, dividing point cloud bounding box to cells, feature extraction and normalization, gesture classification. Modifications of the descriptors are proposed in order to increase hand posture recognition rates and decrease quantity of used features as well as computational cost of the algorithm. The experiments performed on four challenging datasets using cross-validation tests prove the usefulness of our approach
Analysis of Distributed Systems Dynamics with Erlang Performance Lab
Get PDFModern, highly concurrent and large-scale systems require new methods for design, testing and monitoring. Their dynamics and scale require real-time tools, providing a holistic view of the whole system and the ability of showing a more detailed view when needed. Such tools can help identifying the causes of unwanted states, which is hardly possible with static analysis or metrics-based approach. In this paper a new tool for analysis of distributed systems in Erlang is presented. It provides real-time monitoring of system dynamics on different levels of abstraction. The tool has been used for analyzing a large-scale urban traffic simulation system running on a cluster of 20 computing nodes
Multi-objective optimization of vehicle routing problem using evolutionary algorithm with memory
Get PDFThe idea of a new evolutionary algorithm with memory aspect included is proposed to find multiobjective optimized solution of vehicle routing problem with time windows. This algorithm uses population of agents that individually search for optimal solutions. The agent memory incorporates the process of learning from the experience of each individual agent as well as from the experience of the population. This algorithm uses crossover operation to define agents evolution. In the paper we choose as a base the Best Cost Route Crossover (BCRC) operator. This operator is well suited for VPRTW problems. However it does not treat both of parent symmetrically what is not natural for general evolutionary processes. The part of the paper is devoted to find an extension of the BCRC operator in order to improve inheritance of chromosomes from both of parents. Thus, the proposed evolutionary algorithm is implemented with use of two crossover operators: BCRC and its extended-modified version. We analyze the results obtained from both versions applied to Solomon’s and Gehring & Homberger instances. We conclude that the proposed method with modified version of BCRC operator gives statistically better results than those obtained using original BCRC. It seems that evolutionary algorithm with memory and modification of Best Cost Route Crossover Operator lead to very promising results when compared to the ones presented in the literature
One-dimensional fully automatic h-adaptive isogeometric finite element method package
Get PDFThis paper deals with an adaptive finite element method originally developedby Prof. Leszek Demkowicz for hierarchical basis functions. In this paper, weinvestigate the extension of the adaptive algorithm for isogeometric analysisperformed with B-spline basis functions. We restrict ourselves to h-adaptivity,since the polynomial order of approximation must be fixed in the isogeometriccase. The classical variant of the adaptive FEM algorithm, as delivered by thegroup of Prof. Demkowicz, is based on a two-grid paradigm, with coarse andfine grids (the latter utilized as a reference solution). The problem is solved independentlyover a coarse mesh and a fine mesh. The fine-mesh solution is thenutilized as a reference to estimate the relative error of the coarse-mesh solutionand to decide which elements to refine. Prof. Demkowicz uses hierarchicalbasis functions, which (though locally providing C p−1 continuity) ensure onlyC 0 on the interfaces between elements. The CUDA C library described in thispaper switches the basis to B-spline functions and proposes a one-dimensionalisogeometric version of the h-adaptive FEM algorithm to achieve global C p−1continuity of the solution
RESTORING TONE-MARKS IN STANDARD YORÙBÁ ELECTRONIC TEXT: IMPROVED MODEL
Get PDFDiacritic Restoration is a necessity in the processing of languages with Latinbased scripts that utilizes letters outside the basic Latin alphabet used by English language. Yorùbá is one such languages, marking underdot (dot-below)on three characters and tone marks on all seven vowels and two syllabic nasals. The problem of restoring underdotted characters has been fairly addressed using character as linguistic units for restoration. However, the existing characterbased approaches and word-based approach has not been able to sufficiently address restoration of tone marks in Yorùbá. We address in this study tone marks restoration as a subset of diacritic restoration.We proposed using the syllable (derived from word) as the linguistic token for tone marks restoration. In our experimental setup, we used Yoruba text collected from various sources as data with total word count of 250,336 words. These words, on syllabification, yielded 464,274 syllables. The syllables were divided into training and testing data in different proportions ranging from 99% used for training and 1% used for testing to 70% used for training and 30% used for testing. The aim of evaluation different proportions was to determine how the ratio of training-to-test data affect the variations that may occur in the result. We applied Memory-based learning to train the models. We also set up a similar experiment using character token to be able to compare the performance.The result showed that using syllable was able to increase accuracy at word level to 96.23% and an average of almost 15% over that gotten from using character. We also found out that using 75% of data for training and the remaining 25% for testing gives the results with the least variation in a ten-fold cross validation test. Hybridizing the syllable „based approach with other methods like lexicon lookup might likely lead to improvement over the current result
Optimal selection of numerical models for flood embankment pore pressure and temperature data
Get PDFThe aim of the ISMOP project is to study processes in earthen flood embankments: water filtration, pore pressure changes, and temperature changes due to varying water levels in the riverbed. Developing a system for continuous monitoring of flood embankment stability is the main goal of this project. A full-size earthen flood embankment with built-in sensors was built in Czernichow and used to conduct experiments involving the simulation of different flood waves, with parameters mostly measured at time intervals of 15 minutes. Numerical modelling—in addition to providing information about phenomena occurring in the embankment due to external factors and changes in water level—could be used to assess the state of the embankment. Modelling was performed using Itasca Flac 2D 7.0 with an assumed grid cell size of 10x10 cm. The water level in the embankment simulated the water flow in the Wisła River and the temperature of the air and water. Data about the state of the flood embankment was exported every hour.Using numerical models and real experiment data, a model-driven module was used to perform comparisons. Analyses of each half-section of the flood embankment were carried out separately using similarity measures and an aggregate window.For the tests, the North-West (NW) half cross-section of the embankment was chosen, which contains pore pressure and temperature sensors UT6 to UT10. The water level in the embankment was raised to a height of 3m; the best numerical model was considered the one that best matched the actual data recorded by the sensors during the experiment. The experiment period was from 9pm on 29/08/2016 to 9am on 03/09/2016.Seventeen numerical models of the water level rising to 2, 3, and 4 meters were compared against real experimental data from the NW half cross-section. The first step was to verify the similarity between the incoming data from the sensors. If the correlation value exceeded 0.8, the data from the sensors was averaged. The experimental data was then compared against the numerical models using least absolute deviations L1-Norm. The L1-Norm varied from 26 to 32, depending on window length and the numerical model used
Compression of image sequences in interactive medical teleconsultations
Get PDFInteractive medical teleconsultations are an important tool in the modern medical practice. Their applications include remote diagnostics, conferences, workshops and classes for students. In many cases standard medium or low-end machines are employed and the teleconsultation systems must be able to provide high quality of user experience with very limited resources. Particularly problematic are large datasets, consisting of image sequences, which need to be accessed fluently. The main issue is insufficient internal memory, therefore proper compression methods are crucial. However, a scenario where image sequences are kept in a compressed format in the internal memory and decompressed on-the-fly when displayed, is difficult to implement due to performance issues. In this paper we present methods for both lossy and lossless compression of medical image sequences, which require only compatibility with Pixel Shader 2.0 standard, which is present even on relatively old, low-end devices. Based on the evaluation of quality, size reduction and performance, the methods are proved to be suitable and beneficial for the medical teleconsultation applications