Towards Constructing Emotional Landscapes with Music

This chapter describes how the authors arrived at a new paradigm for human-computer interaction that they call tropic mediation. They describe the origins of the research in a wish to provide a concert planner with an expert system. Some consideration is given to how music might have arisen within human culture and, in particular, why it presents unique problems of verbal description. An initial investigation into a discrete, stable lexicon of musical effect is summarized and the authors explain how and why they reached their current work on a computable model of word connotation rather than reference. It is concluded that machines, in order to communicate with people, will need to work with a model of emotional implication to approach the “human” sense of words.

Multilingual Lyric Modeling and Management

This chapter presents a new way to model multilingual lyrics within symbolic music scores. This new model allows one to “plug” onto the symbolic score different lyrics depending on the language. This is done by keeping separate the music notation model and the lyrics model. An object-oriented model of music notation and for lyrics representation are presented with many examples. These models have been implemented in the music editor produced within the WEDELMUSIC IST project. A specific language has been developed to associate the lyrics with the score; the language is able to represent syllables, melismas (extended syllables), refrains, etc. Moreover, the most important music notation formats are reviewed focusing on their representation of multilingual lyrics.

Optical Music Analysis for Printed Music Score and Handwritten Music Manuscript

This chapter describes an optical document imaging system to transform paper-based music scores and manuscripts into machine-readable format and a restoration system to touch-up small imperfections (for example broken stave lines and stems), to restore deteriorated master copy for reprinting. The chapter presents a brief background of this field, discusses the main obstacles, and presents the processes involved for printed music scores processing; using a divide-and-conquer approach to sub-segment compound musical symbols (e.g., chords) and inter-connected groups (e.g., beamed quavers) into lower-level graphical primitives (e.g., lines and ellipses) before recognition and reconstruction. This is followed by discussions on the developments of a handwritten manuscripts prototype with a segmentation approach to separate handwritten musical primitives. Issues and approaches for recognition, reconstruction and revalidation using basic music syntax and high-level domain knowledge, and data representation are also presented.

Lyric Recognition and Christian Music

The chapter is about lyric recognition in Optical Music Recognition (OMR). Discussion is made in the context of Christian music where the lyric is definitive of the genre. Lyrics are obviously found in other music contexts, but they are of primary importance in Christian music — where the words are as integral as the notation. This chapter (i) identifies the inseparability of notation and word in Christian music, (ii) isolates the challenges of lyric recognition in OMR providing some examples of lyric recognition achieved by current scanning software and (iii) considers some solutions outlining page segmentation and character/word recognition approaches, particularly focusing upon the target of recognition, as a high level representation language, that integrates the music with lyrics. The motivation for this chapter includes the observation that high quality lyric recognition is largely omitted by OMR research, but in the context of a music genre inseparable from the word, it is vital. Theoretical, practical (typesetting arrangements/singing synthesis) and philosophical reasons motivate a better examination of lyric recognition.

Modeling Music Notation in the Internet Multimedia Age

Music notation modeling is entering the new multimedia Internet age. In this era new interactive applications are appearing on the market, such as software tools for music tuition and distance learning, for showing historical perspective of music pieces, for musical content fruition in libraries, etc. For these innovative applications several aspects have to be integrated with the model of music notation and several new functionalities have to be implemented, such as automatic formatting, music notation navigation, synchronization of music notation with real audio, etc. In this chapter, the WEDELMUSIC XML format for multimedia music applications of music notation is presented. It includes a music notation format in XML and a format for modeling multimedia element, their relationships and synchronization with a support for digital right management (DRM). In addition, a comparison of this new model with the most important and emerging models is reported. The taxonomy used can be useful for assessing and comparing suitability of music notation models and format for their adoption in new emerging applications and for their usage in classical music editors.

Staff Detection and Removal

This chapter describes the issues involved in the detection and removal of stavelines of musical scores. This removal process is an important step for many Optical Music Recognition systems and facilitates the segmentation and recognition of musical symbols. The process is complicated by the fact that most music symbols are placed on top of stavelines and these lines are often neither straight nor parallel to each other. The challenge here is to remove as much of stavelines as possible while preserving the shapes of the musical symbols, which are superimposed on stavelines. Various problematic examples are illustrated and a detailed explanation of an algorithm is presented. Image processing techniques used in the algorithm include: run-length coding, connected-component analysis, and projections.

An Off-Line Optical Music Sheet Recognition

The optical music recognition is a key problem for coding music sheets of western music in the digital world. The most critical phase of the optical music recognition process is the first analysis of the image sheet. In optical processing of music or documents, the first analysis consists of segmenting the acquired sheet into smaller parts in order to recognize the basic symbols that allow reconstructing the original music symbol. In this chapter, an overview of the main issues and a survey of the main related works are discussed. The O3MR system (Object Oriented Optical Music Recognition) system is also described. The used approach in such system is based on the adoption of projections for the extraction of basic symbols that constitute graphic elements of the music notation. Algorithms and a set of examples are also included to better focus concepts and adopted solutions.

Pen-Based Input for On-Line Handwritten Notation

This chapter is concerned with a novel pen-based interface for (handwritten) music notation. The chapter makes a survey of the current scope of on-line (or dynamic) handwritten input of music notation, presenting the outstanding problems in recognition. A solution using the multi-layer perceptron artificial neural network is presented explaining experiments in music symbol recognition from a study involving notation writing from some 25 people using a pressure-sensitive digitiser for input. Results suggest that a voting system among networks trained to recognize individual symbols produces the best recognition rate in the order of 92% for correctly recognising a positive example of a symbol and 98% in correctly rejecting a negative example of the symbol. A discussion is made of how this approach can be used in an interface for a pen-based music editor. The motivation for this chapter includes (i) the practical need for a pen-based interface capable of recognizing unconstrained handwritten music notation, (ii) the theoretical challenges that such a task presents for pattern recognition and (iii) the outstanding neglect of this topic in both academic and commercial respects.

Wavelets for Dealing with Super-Imposed Objects in Recognition of Music Notation

This chapter is concerned with a problem that arises in Optical Music Recognition (OMR) when notes and other music notation symbols are super-imposed upon stavelines in the music image. We investigate a general-purpose, knowledge-free method of image filtering using two-dimensional wavelets to separate the super-imposed objects. Some background is given to the area of wavelets and a demonstration of how stavelines can be located in a wavelet-filtered image. We also explore the separation of foreground objects (notes) from the background (stavelines) over a variety of image resolutions, in binary and greyscale images using a pixel-based truth representation of the image to evaluate the accuracy with which symbols are identified. We find that the Coifmann family of wavelets appear most suitable for vertical image components, and the Daubechies for the horizontal. The motivation for this chapter stems from the desire to (i) make an original wavelet application in image processing, (ii) provide a fresh (theoretical) perspective on the problem of super-imposed objects in music notation, recognizing the duality of the segregation task that exists with staveline removal/symbol extraction and (iii) evaluate how beneficial wavelet image filtering might be to the OMR process.

Evaluation in the Visual Perception of Music

This chapter is about the need to evaluate the recognition performed by (i) Optical Music Recognition (OMR) systems, and also by (ii) counterpart on-line systems that directly recognize handwritten music input through a pen-based interface. It presents a summary of reviews that have been performed for commercial OMR systems and addresses some of the issues in evaluation that must be taken into account to enable adequate comparison of recognition performance. A representation language [HEART (HiErARchical Text-Based Representation)] is suggested, such that the semantics of music is captured (including the dynamics of handwritten music) and, hence, a target representation provided for recognition processes. Initial consideration of the range of test data that is needed (MusicBase I and II) is also made. The chapter is motivated by the outstanding need for (i) a greater understanding of how to evaluate the accuracy of music recognition systems, (ii) a widely available database of music test data (potentially automatically generated), (iii) an expression of this test data in a format that permits evaluation (for OMR and on-line systems) and (iv) the proliferation of representation languages — none of which captures the dynamics of handwritten music.