| | SLO | ENG | Piškotki in zasebnost

Večja pisava | Manjša pisava

Iskanje po katalogu digitalne knjižnice Pomoč

Iskalni niz: išči po
išči po
išči po
išči po
* po starem in bolonjskem študiju


1 - 10 / 85
Na začetekNa prejšnjo stran123456789Na naslednjo stranNa konec
A VAN-Based Multi-Scale Cross-Attention Mechanism for Skin Lesion Segmentation Network
Shuang Liu, Zeng Zhuang, Yanfeng Zheng, Simon Kolmanič, 2023, izvirni znanstveni članek

Opis: With the rise of deep learning technology, the field of medical image segmentation has undergone rapid development. In recent years, convolutional neural networks (CNNs) have brought many achievements and become the consensus in medical image segmentation tasks. Although many neural networks based on U-shaped structures and methods, such as skip connections have achieved excellent results in medical image segmentation tasks, the properties of convolutional operations limit their ability to effectively learn local and global features. To address this problem, the Transformer from the field of natural language processing (NLP) was introduced to the image segmentation field. Various Transformer-based networks have shown significant performance advantages over mainstream neural networks in different visual tasks, demonstrating the huge potential of Transformers in the field of image segmentation. However, Transformers were originally designed for NLP and ignore the multidimensional nature of images. In the process of operation, they may destroy the 2D structure of the image and cannot effectively capture low-level features. Therefore, we propose a new multi-scale cross-attention method called M-VAN Unet, which is designed based on the Visual Attention Network (VAN) and can effectively learn local and global features. We propose two attention mechanisms, namely MSC-Attention and LKA-Cross-Attention, for capturing low-level features and promoting global information interaction. MSC-Attention is designed for multi-scale channel attention, while LKA-Cross-Attention is a cross-attention mechanism based on the large kernel attention (LKA). Extensive experiments show that our method outperforms current mainstream methods in evaluation metrics such as Dice coefficient and Hausdorff 95 coefficient.
Ključne besede: CNNs, deep learning, medical image processing, NLP, semantic segmentation
Objavljeno v DKUM: 14.03.2024; Ogledov: 401; Prenosov: 298
.pdf Celotno besedilo (1,46 MB)
Gradivo ima več datotek! Več...

Acoustic Gender and Age Classification as an Aid to Human–Computer Interaction in a Smart Home Environment
Damjan Vlaj, Andrej Žgank, 2023, izvirni znanstveni članek

Opis: The advanced smart home environment presents an important trend for the future of human wellbeing. One of the prerequisites for applying its rich functionality is the ability to differentiate between various user categories, such as gender, age, speakers, etc. We propose a model for an efficient acoustic gender and age classification system for human–computer interaction in a smart home. The objective was to improve acoustic classification without using high-complexity feature extraction. This was realized with pitch as an additional feature, combined with additional acoustic modeling approaches. In the first step, the classification is based on Gaussian mixture models. In thesecond step, two new procedures are introduced for gender and age classification. The first is based on the count of the frames with the speaker’s pitch values, and the second is based on the sum of the frames with pitch values belonging to a certain speaker. Since both procedures are based on pitch values, we have proposed a new, effective algorithm for pitch value calculation. In order to improve gender and age classification, we also incorporated speech segmentation with the proposed voice activity detection algorithm. We also propose a procedure that enables the quick adaptation of the classification algorithm to frequent smart home users. The proposed classification model with pitch values has improved the results in comparison with the baseline system.
Ključne besede: acoustic classification, acoustic signal processing, Gaussian mixture model, pitch analysis, smart home
Objavljeno v DKUM: 11.12.2023; Ogledov: 359; Prenosov: 18
.pdf Celotno besedilo (2,07 MB)
Gradivo ima več datotek! Več...

Is the essential facilities doctrine fit for access to data cases? The data protection aspect
Rok Dacar, 2022, izvirni znanstveni članek

Opis: Personal data can be of great economic value for companies as it is an essential input for the offering of a wide array of services. One way for a company to obtain access to essential personal data controlled by another company is by demanding mandatory access on the grounds of the essential facilities doctrine. Such access, however, can violate the right to the protection of personal data of the data subjects if it is not based on one of the legitimate grounds for the processing of personal data set by the GDPR. Two of these grounds are especially likely to be applicable to the access to personal data mandated using the essential facilities doctrine: the interpretation of the Commission decision or the judgment of the Court of Justice ordering the granting of access as a legal obligation and the legitimate interest of the company requesting access, for such access. The anonymisation of personal data is not a viable option for the circumvention of the rules of the GDPR as anonymised personal data loses most of its economic relevance for companies.
Ključne besede: essential facilities doctrine, right to protection of personal data, grounds for processing personal data, anonymisation of personal data, General Data Protection Regulation
Objavljeno v DKUM: 26.09.2023; Ogledov: 281; Prenosov: 8
.pdf Celotno besedilo (657,05 KB)
Gradivo ima več datotek! Več...

Processing of signals produced by strain gauges in testing measurements of the bridges
Boštjan Kovačič, Rok Kamnik, Andrej Štrukelj, Nikolai Ivanovich Vatin, 2015, objavljeni znanstveni prispevek na konferenci

Opis: Practical example of signal processing from strain gauge, inductive transducer and total station measurements are used to illustrate the features of the bridge load testing measurements. FFT provides accurate representation of physical behavior for static and dynamic signals obtained when loading the bridge. As a reference measurement the signal from inductive transducer was taken. A static part of the load test was also geodetically measured and theoretically calculated. The results are comparable.
Ključne besede: bridges, measurements, strain gauge, load test, deformation, strain, signal processing
Objavljeno v DKUM: 12.07.2023; Ogledov: 383; Prenosov: 18
.pdf Celotno besedilo (500,43 KB)
Gradivo ima več datotek! Več...

An overview of selected pseudo-stereophonic techniques : B.Sc. Thesis
Michele Perrone, 2020, diplomsko delo

Opis: Stereophony is a method of sound recording and sound reproduction that uses two separate audio channels, two for each process. In contrast, monophony uses only one audio channel. Stereophonic sound is thus affine, to some degree, to human binaural hearing because it can reproduce some of the ambience and localization cues that were present during the recording session. For this reason, stereophonic music recordings are generally preferred over their monophonic counterparts. To make the sound of monophonic recordings more natural, many pseudo-stereophonic techniques have been developed; these techniques process monophonic sound into stereophonic sound. In this thesis, we review in detail five such techniques, of which we implemented and tested two. Our results show that pseudo-stereophony, when applied appropriately, can create more natural nuances in monophonic recordings and can reduce listening fatigue.
Ključne besede: Pseudo-stereophony, audio processing, music
Objavljeno v DKUM: 14.12.2020; Ogledov: 1091; Prenosov: 65
.pdf Celotno besedilo (752,88 KB)

Signal processor for optical fiber sensors based on MEMS Fabry-Perot interferometer : master's thesis
Nikola Uremović, 2020, magistrsko delo

Opis: In the master thesis, we have created an interrogation system for measuring the change of the optical path in the Fabry-Perot interferometer caused by the strain of the surface at which it was attached to. The change of strain can be calculated via a change in optical path length which is visible as a shift in phase angle. The system will be used as a system for measuring strain, although it can be used for measuring various physical parameters that can cause a change in optical path length, such as pressure, force, temperature, etc. Initially, the theoretical background of the system and project components are represented and explained, following the building process of the electronic and optical part of the project. Lastly, the working principle and programming algorithms of a system are presented and explained. Measurement results are shown at the end, as well as the conclusion that is drawn from the thesis.
Ključne besede: optical fibers, sensors, interferometer, signal processing
Objavljeno v DKUM: 04.11.2020; Ogledov: 1098; Prenosov: 109
.pdf Celotno besedilo (5,11 MB)

Estimating the size of plants by using two parallel views
Barbara Videc, Jurij Rakun, 2017, izvirni znanstveni članek

Opis: This paper presents a method of estimating the size of plants by using two parallel views of the scene, taken by a common digital camera. The approach relays on the principle of similar triangles with the following constraints: the resolution of the camera is known; the object is always in parallel to the camera sensor and the intermediate distance between the two concessive images is available. The approach was first calibrated and tested using one artificial object in a controlled environment. After that real examples were taken from agriculture, where we measured the distance and the size of a vine plant, apple and pear tree. By comparing the calculated values to measured values, we concluded that the average absolute error in distance was 0.11 m or around 3.7 %, and the absolute error in high was 0.09 m or 4.6 %.
Ključne besede: digital image processing, size, digital camera, pixels, similar triangles
Objavljeno v DKUM: 10.10.2018; Ogledov: 1226; Prenosov: 304
.pdf Celotno besedilo (727,92 KB)
Gradivo ima več datotek! Več...

Determining the grain size distribution of granular soils using image analysis
Nihat Dipova, 2017, izvirni znanstveni članek

Opis: Image-processing technology includes storing the images of objects in a computer and processing them with the computer for a specified purpose. Image analysis is the numerical expression of the images of objects by means of mimicking the functioning of the human visual system and the generation of numerical data for calculations that will be made later. Digital image analysis provides the capability for rapid measurement, which can be made in near-real time, for numerous engineering parameters of materials. Recently, image analysis has been used in geotechnical engineering practices. Grain size distribution and grain shape are the most fundamental properties used to interpret the origin and behaviour of soils. Mechanical sieving has some limitations, e.g., it does not measure the axial dimension of a particle, particle shape is not taken into consideration, and especially for elongated and flat particles a sieve analysis will not yield a reliable measure. In this study the grain size distribution of sands has been determined following image-analysis techniques, using simple apparatus, non-professional cameras and open-code software. The sample is put on a transparent plate that is illuminated with a white backlight. The digital images were acquired with a CCD DSLR camera. The segmentation of the particles is achieved by image thresholding, binary coding and particle labeling. The geometrical measurements of each particle are obtained using an automated pixel-counting technique. Local contacts or limited overlaps were overcome using a watershed split. The same sample was tested by traditional sieve analysis. An image-analysis-based grain size distribution has been compared with a sieve-analysis distribution. The results show that the grain size distribution of the image-based analysis and the sieve analysis are in good agreement.
Ključne besede: image analysis, image processing, grain size, sand
Objavljeno v DKUM: 18.06.2018; Ogledov: 1377; Prenosov: 149
.pdf Celotno besedilo (1,27 MB)
Gradivo ima več datotek! Več...

Processing methodology and dialectological aspects of the Dictionary of Moravian and Silesian anoikonyms (minor place names)
Libuše Čižmárová, 2010, izvirni znanstveni članek

Opis: This paper presents the routines used by Brno linguists working on the Dictionary of Moravian and Silesian Anoikonyms (preparing collective entries introduced by abstract headwords). The output will be primarily a multifunctional interactive digital dictionary. Since most of the material has been recorded in dialect form, the authors must be experienced in dialectology. The computer program offers the possibility to generate maps enabling comparison with dialectological maps of the Czech Linguistic Atlas.
Ključne besede: anoikonyma, onomastics, toponomastics, dictionaries, digitization, computer data processing, Moravia, Silesia, Czech Republic, dialectology, geolinguistics
Objavljeno v DKUM: 05.02.2018; Ogledov: 966; Prenosov: 360
.pdf Celotno besedilo (406,01 KB)
Gradivo ima več datotek! Več...

Dialect materials in the Estonian etymological dictionary
Iris Metsmägi, 2010, izvirni znanstveni članek

Opis: Estonian Etymological Dictionary being compiled at the Institute of the Estonian Language. A limited number of dialect words will be included in the headword list of the dictionary. Dialect data may be vital for etymologization as well. In the case of genuine words and older loanwords, the archaic phonological traits that have survived in dialects are essential; in the case of more recent loanwords, the dialectal variants show different degrees of adaptation. Sometimes the areal distribution of a word may prove a valuable cue to its original background. Etymologization may also be based on the dialectal meaning of a word. A specific group consists of dialect words adopted into standard usage in a different sense as technical terms.
Ključne besede: Estonian, etymological dictionaries, computer data processing, dialectology, linguistic geography
Objavljeno v DKUM: 02.02.2018; Ogledov: 1173; Prenosov: 335
.pdf Celotno besedilo (313,02 KB)
Gradivo ima več datotek! Več...

Iskanje izvedeno v 5.8 sek.
Na vrh
Logotipi partnerjev Univerza v Mariboru Univerza v Ljubljani Univerza na Primorskem Univerza v Novi Gorici