Marcin Korytkowski, Roman Senkerik, Magdalena M. Scherer, Rafal A. Angryk, Miroslaw Kordos and Agnieszka Siwocha
Fast content-based image retrieval is still a challenge for computer systems. We present a novel method aimed at classifying images by fuzzy rules and local image features. The fuzzy rule base is generated in the first stage by a boosting procedure. Boosting meta-learning is used to find the most representative local features. We briefly explore the utilization of metaheuristic algorithms for the various tasks of fuzzy systems optimization. We also provide a comprehensive description of the current best-performing DISH algorithm, which represents a powerful version of the differential evolution algorithm with effective embedded mechanisms for stronger exploration and preservation of the population diversity, designed for higher dimensional and complex optimization tasks. The algorithm is used to fine-tune the fuzzy rule base. The fuzzy rules can also be used to create a database index to retrieve images similar to the query image fast. The proposed approach is tested on a state-of-the-art image dataset and compared with the bag-of-features image representation model combined with the Support Vector Machine classification. The novel method gives a better classification accuracy, and the time of the training and testing process is significantly shorter.
Marcin Zalasiński, Krystian Łapa, Krzysztof Cpałka, Krzysztof Przybyszewski and Gary G. Yen
The on-line signature is a biometric attribute which can be used for identity verification. It is a very useful characteristic because it is commonly accepted in societies across the world. However, the verification process using this particular biometric feature is a rather difficult one. Researchers working on identity verification involving the on-line signature might face various problems, including the different discriminative power of signature descriptors, the problem of a large number of descriptors, the problem of descriptor generation, etc. However, population-based algorithms (PBAs) can prove very useful when resolving these problems. Hence, we propose a new method for on-line signature partitioning using a PBA in order to improve the verification process effectiveness. Our method uses the Differential Evolution algorithm with a properly defined evaluation function for creating the most characteristic partitions of the dynamic signature. We present simulation results of the proposed method for the BioSecure DS2 database distributed by the BioSecure Association.
Piotr Duda, Maciej Jaworski, Andrzej Cader and Lipo Wang
In recent years, many deep learning methods, allowed for a significant improvement of systems based on artificial intelligence methods. Their effectiveness results from an ability to analyze large labeled datasets. The price for such high accuracy is the long training time, necessary to process such large amounts of data. On the other hand, along with the increase in the number of collected data, the field of data stream analysis was developed. It enables to process data immediately, with no need to store them. In this work, we decided to take advantage of the benefits of data streaming in order to accelerate the training of deep neural networks. The work includes an analysis of two approaches to network learning, presented on the background of traditional stochastic and batch-based methods.
Robert K. Nowicki, Konrad Grzanek and Yoichi Hayashi
The paper presents the idea of connecting the concepts of the Vapnik’s support vector machine with Pawlak’s rough sets in one classification scheme. The hybrid system will be applied to classifying data in the form of intervals and with missing values . Both situations will be treated as a cause of dividing input space into equivalence classes. Then, the SVM procedure will lead to a classification of input data into rough sets of the desired classes, i.e. to their positive, boundary or negative regions. Such a form of answer is also called a three–way decision. The proposed solution will be tested using several popular benchmarks.
Ratnesh Kumar, Edwin Weill, Farzin Aghdasi and Parthasarathy Sriram
In this paper we tackle the problem of vehicle re-identification in a camera network utilizing triplet embeddings. Re-identification is the problem of matching appearances of objects across different cameras. With the proliferation of surveillance cameras enabling smart and safer cities, there is an ever-increasing need to re-identify vehicles across cameras. Typical challenges arising in smart city scenarios include variations of viewpoints, illumination and self occlusions. Most successful approaches for re-identification involve (deep) learning an embedding space such that the vehicles of same identities are projected closer to one another, compared to the vehicles representing different identities. Popular loss functions for learning an embedding (space) include contrastive or triplet loss. In this paper we provide an extensive evaluation of triplet loss applied to vehicle re-identification and demonstrate that using the recently proposed sampling approaches for mining informative data points outperform most of the existing state-of-the-art approaches for vehicle re-identification. Compared to most existing state-of-the-art approaches, our approach is simpler and more straightforward for training utilizing only identity-level annotations, along with one of the smallest published embedding dimensions for efficient inference. Furthermore in this work we introduce a formal evaluation of a triplet sampling variant (batch sample) into the re-identification literature. In addition to the conference version , this submission adds extensive experiments on new released datasets, cross domain evaluations and ablation studies.
The paper presents the use of Taguchi method to optimize the cutting of stainless steel by Abrasive Water Jet. Shown are the influence of the most important machining parameters, such a traverse speed, abrasive grains size and concentration of abrasive in the jet on the maximum depth of cut. Analysis of variance - ANOVA was used to determine the effect of machining parameters on the cutting depth. Based on the calculated signal/noise ratios for individual parameters of the cutting process, their impact on cutting depth was determined and optimal process conditions were determined in order to reach the maximum depth of cut. The empirical verification of this process was also performed by comparing the depth of cut predicted and achieved in the tests.
The research was aimed at analysing the factors that affect the accuracy of merging point clouds when scanning over longer distances. Research takes into account the limited possibilities of target placement occurring while scanning opposite benches of quarries or open-pit mines, embankments from opposite banks of rivers etc. In all these cases, there is an obstacle/void between the scanner and measured object that prevents the optimal location of targets and enlarging scanning distances. The accuracy factors for cloud merging are: the placement of targets relative to the scanner and measured object, the target type and instrument range. Tests demonstrated that for scanning of objects with lower accuracy requirements, over long distances, it is optimal to choose flat targets for registration. For objects with higher accuracy requirements, scanned from shorter distances, it is worth selecting spherical targets. Targets and scanned object should be on the same side of the void.
Based on Gauss’ law for the electric field, new formulas were deduced, that enable for the first time the writing of an analytical formula of the built-in potential of implanted and diffused semiconductor junctions. Consequently, in this work is devised a measurement technique for the built-in potential of such junctions. Such measurement is useful because new semiconductor materials besides silicon are more and more used today, like silicon-carbide (SiC) and gallium-nitride (GaN), which have larger bandgap and junction built-in potential. Finding the built-in potential helps adjusting the computer assisted design (CAD) tools and validates the simulation of such wide-bandgap devices.
The study defines the concept of e-larning and presents the requirements for conducting an effective course to achieve the assumed learning outcomes. As an element of the introduction to the subject, the advantages and disadvantages of e-learning resulting from the experience over the years of organizations that used or continue to use such forms of teaching are presented. The research part presents the results of research carried out in a deliberately selected group of academic teachers and a group of students. The research results presented and discussed include such aspects as: expectations, advantages, disadvantages, fit, convergence of opinions and proposals for improvement of classes conducted in the e-learning mode in the field of Safety and Health at Work. The study is completed by conclusions and proposals of utilitarian solutions in the field of conducting academic education for selected subjects in a fixed time dimension in the form of e-learning.
Organic dyes are widely used in many industries (textiles, food, cosmetics, medicine and biology). These plants produce wastewater containing dyes. Even small amounts of dyes can cause a strong colour of wastewater. Therefore, it is very important to effectively remove residues of these pollutants from the wastewater, before discharging them into the environment. The sorption process is one of the methods used to remove dyes. However this method is often unprofitable economically in comparison with other dye removal processes, due to the high cost of commercial sorbents. Therefore, research is currently conducted in order to find waste materials that can be used as sorbents. The static sorption process of two dyes were carried out (Acid Red 18 and Acid Green 16) with the use of dried excess sludge. The activated sludge (excess) came from a municipal sewage treatment plant that purifies wastewater from carbon, nitrogen and phosphorus compounds. During the study the most favourable pH of the process and the contact time of the sorbent with the dyes were determined. It was observed that for both dyes the highest effectiveness of dye removal was obtained at pH = 2. The most favourable contact time was equal to 60 and 180 minutes for Acid Red 18 and Acid Green 16 respectively. In addition, in order to establish process parameters, a different models of sorption isotherm was examined. The studies showed that the sorption capacity (calculated based on Langmuir model) was much higher in the case of Acid Green 16 (qm = 434.8 mg/g) than for Acid Red 18 (qm = 109.9 mg/g). The experiments to evaluate the effect of pH, contact time on the process effectiveness and to determine the sorption isotherm were conducted at 293.15 K.