Matevž Pesek

University of Ljubljana · Faculty of Computer and Information Science

This chapter presents our findings on emotional and color perception of music. It emphasizes the importance of user-aware music information retrieval (MIR) and the advantages that research on emotional processing and interaction between multiple modalities brings to the understanding of music and its users. Analyses of results show that correlation...

We present a compositional hierarchical model for symbolic music representations (SymCHM). The SymCHM is a deep architecture model with a transparent multi-layer structure. The model can be used for pattern discovery, as well as music music similarity. It can learn a set of representative repeated patterns of individual works or larger corpora in a...

Advances in multimedia and information systems have shifted the focus from general content repositories towards personalized systems. Much effort has been put into modeling and integration of affective states with the purpose of improving overall user experience and functionality of the system. In this chapter, we present a multi-modal dataset of u...

The paper presents a new compositional hierarchical model for robust music transcription. Its main features are unsupervised learning of a hierarchical representation of input data, transparency, which enables insights into the learned representation, as well as robustness and speed which make it suitable for real-world and real-time use. The model...

Troubadour platform is an open-source personalized and adaptive web platform for ear training. The platform was developed to support music theory classes with automated music-theory-related exercises. In this paper, we present our three-stage development methodology, which incorporated the needs and feedback from both teachers and students to build...

V zadnjem desetletju je zanimanje za kriptovalute močno naraslo. S tem je prišlo tudi do porasta kapitala, zaklenjenega v pametnih pogodbah, kar pa je hkrati povečalo tveganje za napade. Za preprečevanje napadov na pametne pogodbe je potrebna analiza programerskih vzorcev in okolja, v katerem se pametne pogodbe izvajajo. V prispevku obravnavamo var...

Nove funkcionalnosti na mobilni platformi za učenje glasbene teorije so usmerjene v prima vista izvajanje ritmičnih in melodičnih vzorcev ter so integrirane v celostno zasnovo platforme, saj ohranjajo enotne poigritvene elemente. Evalvacija uporabniške izkušnje novih igrificiranih pripomočkov je ponudila pozitivne povratne informacije, ki so spodbu...

When implementing cryptographic primitives in software, we often resort to optimizing the software code in order to reduce unnecessary computation time. Such decisions can have negative consequences on data leakage outside the implementation. Attackers can exploit the additional information generated by the execution of the software solution in a w...

V kontekstu povezovanja različnih informacijskih sistemov je razreševanje domenskih naslovov ključni proces identifikacije deležnikov v širšem okolju IT infrastrukture, ki ob pomanjkljivi konfiguraciji lahko predstavlja tveganje za zlorabo s strani napadalcev. Zaradi rastoče kompleksnosti infrastrukture se količina takšnih vektorjev napada na infor...

V računalništvu je upravljanje pomnilnika ključnega pomena za delovanje programske opreme. Napake pri upravljanju pomnilnika lahko privedejo do resnih varnostnih ranljivosti, ki jih napadalci lahko izkoristijo za izvajanje poljubne kode ali pridobivanje občutljivih podatkov. Ena izmed pogostih napak pri upravljanju pomnilnika je dvojno sproščanje p...

Modern music streaming services rely on recommender systems to help users navigate within their large collections. Collaborative filtering (CF) methods, that leverage past user–item interactions, have been most successful, but have various limitations, like performing poorly among sparsely connected items. Conversely, content-based models circumven...

Članek predstavlja pregled metodologij za kvalitativno vrednotenje odprtih podatkov. Različni pristopi so bili razviti tako za ocenjevanje metapodatkov, ki opisujejo posamezne zbirke podatkov, kot tudi za ocenjevanje samih podatkov. Metodologije pogosto temeljijo na standardih in širše sprejetih smernicah za zagotavljanje kakovosti podatkov, posame...

The open data market is growing, driven by laws requiring governments to make data publicly available. However, the true value of this market lies in the products and services derived from open data. In Slovenia, the open data market has been growing since 2015, especially in the field of mobility. Platforms such as Avtolog.si facilitate the retrie...

This study analyzes the effect of using a virtual reality (VR) game as a complementary tool to improve users’ rhythmic performance and perception in a remote and self-learning environment. In recent years, remote learning has gained importance due to various everyday situations; however, the effects of using VR in such situations for individual and...

V luči trenutne okoljske in energetske krize je uporaba javnega in nizkoogljičnega prevoza ključnega pomena za dosego zadanih državnih in evropskih ciljev do konca tekočega desetletja. Za razliko od osebnega avtomobila pa potovanje z drugimi načini prevoza zahteva vsaj nekaj načrtovanja, ki se lahko izkaže za preveliko prilagoditev za marsikaterega...

This study analyzes the effect of using a virtual reality (VR) game as a complementary tool to improve users' rhythmic performance and perception in a remote and self-learning environment. In recent years, remote learning has gained importance due to various everyday situations; however, the effects of using VR in such situations for individual and...

The VR interfaces have become more affordable than ever before. Because of their multimodal interaction capabilities, these interfaces provide a medium to virtually participate in various learning processes that can engage both sensory and motor senses. In this paper, we present an ongoing study to evaluate a VR game designed to enhance four aspect...

Procese razumevanja glasbe kot niza pojavov, ki so tesno povezani z IT praksami iskanja glasbe v slovenski raziskovalni skupnosti, skiciramo s treh osnovnih vidikov: etnomuzikološkega, bibliotekarskega (bistvenega, ko se glasbi približamo računalniško) in IT. Članek ocenjuje doprinos teh perspektiv k razumevanju glasbe, in predlaga, da tri obravnav...

Trubadur je odprtokodna platforma za urjenje glasbenega posluha z avtomatiziranimi vajami ritmičnega in intervalnega nareka. Platformo smo ovrednotili z dijaki Konservatorija za glasbo in balet Ljubljana v šolskih letih 2018/19–2020/21. Rezultati evalvacije so pokazali, da lahko uporaba platforme poveča uspešnost pri testih in predstavlja dopolnite...

This paper presents a new version and a three-month evaluation of the Troubadour platform—an open-source music theory ear training platform. Through interviews with teachers, we gathered the most-needed features which would aid their use of the platform. In the new version of the Troubadour platform, we implemented different types of interaction, i...

Self-regulated learning (SRL) plays an important role in successful learning with hypertext. The use of appropriate SRL strategies helps students acquire new knowledge more efficiently. We investigated the use of SRL strategies in individual learning from expository science hypertext, the correlations between different measures of self-regulation,...

The importance of self-regulated learning (SRL) has increased during the COVID-19 pandemic and measures for assessing students’ self-regulation skills and knowledge are greatly needed. We present the results of the first thorough adaptation of the Children’s Perceived use of Self-Regulated Learning Inventory (CP-SRLI). The inventory, consisting of...

The study of inter-annotator agreement in musical pattern annotations has gained increased attention over the past few years. While expert annotations are often taken as the reference for evaluating pattern discovery algorithms, relying on just one reference is not usually sufficient to capture the complex musical relations between patterns. In thi...

E-gradiva in e-učbeniki si počasi utirajo pot v naše šole. Njihov pravi pomen se je pokazal tudi ob zadnji epidemiji, ko so bile šole zaprte, učenci pa večkrat prepuščeni sami sebi in angažiranosti svojih staršev. Za uspešno učenje z uporabo e-gradiv mora učenec uporabiti primerne strategije samoregulacije učenja, ki mu omogočajo, da doseže zastavl...

E-gradiva in e-učbeniki si počasi utirajo pot v naše šole. Njihov pravi pomen se je pokazal tudi ob zadnji epidemiji, ko so bile šole zaprte, učenci pa večkrat prepuščeni sami sebi in angažiranosti svojih staršev. Za uspešno učenje z uporabo e-gradiv mora učenec uporabiti primerne strategije samoregulacije učenja, ki mu omogočajo, da doseže zastavl...

Herein, we report a computational algorithm that follows a spectroscopist-driven elucidation process of the structure of an organic molecule based on IR, 1 H and 13 C NMR, and MS tabular data. The algorithm is independent from database searching and is based on a bottom-up approach, building the molecular structure from small structural fragments v...

Melody prediction is an important aspect of music listening. The success of prediction, i.e., whether the next note played in a song is the same as the one predicted by the listener, depends on various factors. In the paper, we present two studies, where we assess how music familiarity and music expertise influence melody prediction in human listen...

We present the development and evaluation of a gamified rhythmic dictation application for music theory learning. The application's focus is on mobile accessibility and user experience, so it includes intuitive controls for input of rhythmic exercises, a responsive user interface, several gamification elements and a flexible exercise generator. We...

We present the development and evaluation of a gamified rhythmic dictation application for music theory learning. The application was developed as a web application in the existing Troubadour platform for music ear training. The focus of the developed application was on user experience and engagement. With regards to the former, we developed a resp...

This paper presents a model capable of learning the rhythmic characteristics of a music signal through unsupervised learning. The model learns a multi-layer hierarchy of rhythmic patterns ranging from simple structures on lower layers to more complex patterns on higher layers. The learned hierarchy is fully transparent, which enables observation an...

S povečano popularnostjo globokih arhitektur, ki temeljijo na nevronskih mrežah, so se v zadnjem času bistveno izboljšali rezultati pri reševanju problemov na več področjih. Zaradi popularnosti in uspešnosti teh globokih pristopov, temelječih na nevronskih mrežah, so bili drugi simbolni in kompozicionalni pristopi odmaknjeni od središča pozornosti...

Users of a recommender system may be requested to express their preferences about items either with evaluations of items (e.g. a rating) or with comparisons of item pairs. In this work we focus on the acquisition of pairwise preferences in the music domain. Asking the user to explicitly compare music, i.e., which, among two listened tracks, is pref...

The article presents a method for segmentation of ethnomusicological field recordings. Field recordings are integral documents of folk music performances captured in the field, and typically contain performances, intertwined with interviews and commentaries. As these are live recordings, captured in non-ideal conditions, they usually contain signif...

The article presents an experimental study of the emotional and color perception of music, focusing on folk music in comparison with other musical genres. The analysis showed that some musical emotions are conveyed differently between different genres. Overall, Folk genre had the most balanced distribution of the emotions used in the ratings, along...

In recent years, deep architectures, most commonly based on neural networks, have advanced the state of the art in many research areas. Due to the popularity and the success of deep neural-networks, other deep architectures, including compositional models, have been put aside from mainstream research. This dissertation presents the compositional h...

This paper presents a compositional hierarchical model for pattern discovery in symbolic music. The model can be regarded as a deep architecture with a transparent structure. It can learn a set of repeated patterns within individual works or larger corpora in an unsupervised manner, relying on statistics of pattern occurrences, and robustly infer t...

We present a research tool for user preference elicitation that collects both explicit user feedback and unobtrusively acquired facial expressions. The concrete implementation is a web-based user interface where the user is presented with two music excerpts. After listening to both, the user provides a pairwise score (i.e. which of the two items is...

This paper presents a new multimodal dataset Moodo that can aid the development of affective music information retrieval systems. Moodo’s main novelties are a multimodal approach that links emotional and color perception to music and the inclusion of user context. Analysis of the dataset reveals notable differences in emotion-color associations and...

The paper introduces Stripe, an interactive continuous scale for online surveys that makes it easy to compare multiple answers on a single screen. The Stripe is evaluated as an alternative to the n-point Likert scale, which is commonly used in online usability questionnaires like the System Us-ability Scale (SUS). The paper presents the results of...

This paper presents our compositional hierarchical model for pattern discovery in symbolic music SymCHM and its extension SymCH-MMerge. The model uses unsupervised learning to build a hierarchical structure reflecting the patterns present in the input music representation. It can generate several competing hypotheses on the found patterns, so a pat...

This submission is an extension to the SymCHM approach, submitted to Mirex 2015 Discovery of Repeated Themes & Sections. While retaining the compositional approach, we further improved the results by refining the output the SymCHM, using a technique of pattern merging and output refinement.

The compositional hierarchical model has been well explored for several tasks in the field of music information retrieval, including the automated chord estimation and multiple fundamental frequency estimation. This submission introduces the compositional approach to the symbolic representations of the music scores by using the model’s structure wh...

We present the WebCHM — a web based tool for music analysis, transcription and annotation. The tool is based on our compositional hierarchical model (CHM) for music modeling [1] and has several goals: to provide a free and accessible way for researchers to evaluate the CHM on their datasets; to provide a web-based service for MIR-related music anal...

The educational and IT communities have produced a number of e-learning products, ranging from support-oriented platforms for online courses and learning to educational games. However, there is still a growing need for inclusive and accessible learning products. To meet the need, we developed an accessible online web and mobile platform for educati...

We present an innovative dynamic visual interface, the Mood-Stripe, which provides a continuous-scale, multi-parameter drag-and-drop alternative to the standard n-degree (Likert) scale widgets, commonly used in online evaluation processes. We elaborate on the motivation for the development of the new user input interfaces, and present the results o...

Several studies dealing with music recommendation and visualization base their approaches on datasets gathered with user surveys. However, the gathering procedure is seldom the focus of music research, even though the user interfaces and methodology are an important part of gathering the music data and evaluation of the music information retrieval...

This paper presents a biologically-inspired compositional hierarchical model for MIR. The model can be treated as a deep learning model, and poses an alternative to deep architectures based on neural networks. Its main features are generativeness and transparency that allow clear in-sight into concepts learned from the input music signals. The mode...

The paper presents a new dataset that captures the effect of mood on visual and auditory perception of music. With an online survey, we have collected a dataset of over 6600 responses capturing users' mood, emotions evoked and expressed by music and the perception of color with regard to emotions and music. We describe the methodology of gathering...

The paper addresses the task of automatic audio chord estimation using stacked generalization of multiple classifiers over Hidden Markov model (HMM) estimators. We evaluated two feature types for chord estimation: a new compositional hierarchical model and standard chroma feature vectors. The compositional hierarchical model is presented as an alte...

In this paper we present the Edoo initiative that aims for increased involvement of teachers in the process of e-learning materials creation. Its main goal is to connect two distinct communities-teachers as users of e-learning materials and programmers as creators of e-learning materials-through the online portal and learning site designed for supp...

Chroma feature vectors are widely used for automated chord estimation in the music information retrieval (MIR) �eld. We propose a compositional hierarchical model for unsupervised feature learning providing an alternative to chroma features. Both feature types are further modelled with two separate Hidden Markov models (HMMs) in order to estimate t...