Catálogo de publicaciones - libros

MX is a new XML-based format to describe comprehensively heterogeneous music contents. In a single MX file, music symbols, printed scores, audio tracks, computer-driven performances, catalogue metadata, text and graphic contents related to a single music piece are linked and mutually synchronised within the same framework. Heterogeneous contents are organised in a multilayered structure that supports different encoding formats and a number of digital objects for each layer.

This work describes algorithms dedicated to score and audio alignment using the MX / IEEE P1599 format. The format allows description of the score, and management of synchronisation points, linking them with different versions of the performed music. An algorithm is proposed here that allows alignment of an MX score and its execution, coded in PCM format, which produces an output for the MX Spine that contains synchronisation between notes and audio signal. The proposed architecture is based on two different steps: the first deals with the audio level and the extraction of features like pitch, onset and the like, while the second determines the alignment between the features extracted in the first step and the notes present in the MX score. For each step, different algorithms are proposed and discussed, and analysis and comparison of synchronisation capabilities are provided.

This article presents an approach to music interaction and creativity at the and . While common music interaction occurs at the symbolic level, on a sequence of notes, it is the structural level, the focus of this writing, that allows interaction with music contents at a higher degree. This requires identification and creation of in existing and new compositions, then the resulting set of music themes can be treated with Petri Nets, even in real time. Algorithms can be applied to modify the music synthesised or played, and, thanks to MX, this interaction is performed simultaneously at different levels. After an introduction of the main concepts, the article describes a case study based on “Peaches en Regalia”, by Frank Zappa, presented at the , , which took place in Rome on 9-10 June 2006. Finally, advanced MX scenarios will be presented.

The possibility of interconnecting with allows creation and manipulation of timbres in an exhaustive way. Thanks to Csound features, it is possible to let MX files sound with any synthesis algorithm. This approach supports all the connections to current files through MIDI, real-time controls, analysis and synthesis thanks to the direct interaction with files linked in MX.

In the last few years, the volume of multimedia content available on the Web significantly increased. This led to the need for techniques to handle such data. In this context, we see a growing interest in considering the Semantic Web in action and in the definition of tools capable of analyzing and organizing digital shape models. In this paper, we present a Semantic Web environment, , for inspecting 3D shapes and for structuring and annotating such shapes according to ontology-driven metadata. Specifically, we describe in details the first module of , the and the algorithms we have developed for extracting geometrical and topological information from 3D shapes. We also describe the second module, the which produces a graph-based representation of the decomposition of the shape into manifold components. This is successively modified by the third and the fourth modules, which perform the automatic and manual segmentation of the manifold parts.

We provide an integrated ontological framework offering coverage for deep semantic content, including ontological representation of multimedia based on the MPEG-7 standard. We link the deep semantic level with the media-specific semantic level to operationalize multimedia information. Through the link between multimedia representation and the semantics of specific domains we approach the Semantic Gap. The focus of the paper is on the linguistic features of multimedia, the annotation of these features and their analysis.

This paper presents an approach to semi-automate photo annotation. Instead of using content-recognition techniques this approach leverages context information available at the scene of the photo such as time and location in combination with existing photo annotations to provide suggestions to the user. An algorithm exploits a number of technologies including Global Positioning System (GPS), Semantic Web, Web services and Online Social Networks, considering all information and making a best-effort attempt to suggest both people and places depicted in the photo. The user then selects which of the suggestions are correct to annotate the photo. This process accelerates the photo annotation process dramatically which in turn aids photo search for a wide range of query tools that currently trawl the millions of photos on the Web.

Frame selection techniques based on semantic analysis are widely used to provide content-based summaries of video sequences. Video adaptation is also required to enable the consumption of these summaries through an increasing variety of devices and networks. Scalable video coding provides embedded coding of video where video adaptation is very simple and efficient. In this paper, we describe a model where both video summarization and adaptation are integrated within the MPEG-21 DIA framework, along with some use cases.

The increasing interest in Cognitive Vision Systems (CVS) motivates the apparition of ad-hoc stages designed for the integration of multiple kinds of knowledge. This paper proposes a novel ontology to restrict and integrate high-level semantics for Human Sequence Evaluation (HSE), which targets multilingual capabilities and multipurpose end-user interfaces. The main contributions of this paper are the conception of a neutral semantic layer, which allows to link vision and linguistic domains; and the use of instead of verbs as basic elements for an ontological categorization of occurrences. In our approach, the domain has been restricted to outdoor surveilled scenarios, involving interactions among pedestrians, static objects, and vehicular traffic.

A novel approach to model the semantic knowledge associated to objects detected in images is presented. The model is aimed at the classification of such objects according to contextual information combined to the extracted features. The system is based on Conditional Random Fields, a probabilistic graphical model used to model the conditional a-posteriori probability of the object classes, thus avoiding problems related to source modelling and features independence constraints. The novelty of the approach is in the addressing of the high-level, semantically rich objects interrelationships among image parts. This paper presents the application of the model to this new problem class and a first implementation of the system.