%+ audio analysis (via TimeSide)\r
%+ time markers for annotation and segmentation of instant or temporal region of the audio data.\r
\r
-The main features of \emph{Telemeta} are :\r
+Regarding web aspects, the main features of \emph{Telemeta} are :\r
\begin{itemize}\r
-\item Web platform :\r
- \begin{itemize}\r
- \item \emph{Pure HTML} web user interface including dynamical forms\r
- \item High level \emph{search engine}\r
- \item Smart \emph{workflow management} with contextual user lists, profiles and rights\r
- \item \emph{Multi-language support} (now english and french, german and\r
- spanish in development)\r
- \item RSS and JSON feed generators\r
- \item XML serialized backup\r
- \item Strong SQL or Oracle backend\r
- \item MVC architecture \r
- \end{itemize}\r
-\item Media support :\r
- \begin{itemize}\r
- \item Smart dynamical and skinnable \emph{audio player}\r
-% (thanks to TimeSide and SoundManager2)\r
- \item \emph{Multi-format support} : read all available audio and video formats \r
- \item \emph{Playlist management} for all users with CSV data export\r
- \item On the fly \emph{audio transcoding, analyzing, visualizing and metadata embedding} based on an easy plugin architecture (see Section~\ref{sec:Timeside})\r
- \end{itemize}\r
-\item Metadata : \r
- \begin{itemize}\r
- \item Social cumulative indexing with \emph{semantic ontologies} and\r
- \emph{time-coded markers}\r
- \item \emph{Geo-Navigator} for audio geolocalization\r
- \item DublinCore compatibility and OAI-PMH data provider\r
- \end{itemize}\r
+\item \emph{Pure HTML} web user interface including high level \emph{search engine}\r
- \item \emph{User management} with individual desk, lists, profiles and\r
- rights\r
++\item Smart \emph{workflow management} with contextual user lists, profiles and rights\r
++ % \item RSS and JSON feed generators\r
++ % \item XML serialized backup\r
+\item Strong SQL or Oracle backend\r
++\item MVC architecture \r
\end{itemize}\r
\r
++ \r
++\r
+Beside database management, the audio support is mainly provided through an external component : TimeSide which is described in Section~\ref{sec:Timeside}\r
\r
\subsection{Metadata}\label{sec:metadata}\r
-Beside the audio data, an efficient and dynamic management of the associated metadata is also required. Consulting metadata provide both an exhaustive access to valuable information about the source of the data and to the related work of peer researchers. \r
+In addition to the audio data, an efficient and dynamic management of the associated metadata is also required. %Consulting metadata provide both an exhaustive access to valuable information about the source of the data and to the related work of peer researchers. \r
Dynamically handling metadata in a collaborative manner enable to optimize the continuous process of knowledge gathering and enrichment of the materials in the database. \r
-\r
-One of the major challenge is thus the standardization of audio and metadata formats with the aim of long-term preservation and usage of the different materials. \r
+%One of the major challenge is thus the standardization of audio and metadata formats with the aim of long-term preservation and usage of the different materials.\r
+The compatibility with other systems is facilitated by the integration of the metadata standards protocols \emph{Dublin Core} and \emph{OAI-PMH} \cite{DublinCore,OAI-PMH}.\r
\r
Metadata provide two different kinds of information about the audio item : contextual information and annotations.\r
-Contextual information consists in :\r
-\begin{itemize}\r
-\item Geographic and cultural information (Location details, population/social group, ethnographic context)\r
-\item Musical informations (style, composition, interprets, ...)\r
-\item Archiving data (code and reference to the item)\r
-\item Technical data (media type and duration)\r
-\item Related media (any other material (images, video or text document associated with the audio item)\r
-\end{itemize}\r
+ \r
-Annotation information provides additional comments or analysis done by some expert on the data. Annotations can consist in temporal information such as :\r
-\begin{itemize}\r
-\item segmentation in relevant class or label for ethnomusicological study (e.g. speech versus singing voice segment)\r
-\item time-coded makers for instantaneous comments \r
-\end{itemize}\r
-It should be notice that those annotations can be done either by an human expert or by some audio processing automatic analysis (see Section~\ref{sec:Timeside}).\r
+\paragraph{Contextual Information}\r
- Regarding ethnomusicology, contextual information could be geographic and cultural, musical. It could also store archives related information and include related materials in any multimedia format. \r
++Regarding ethnomusicology, contextual information could be geographic, cultural and musical. It could also store archives related information and include related materials in any multimedia format. \r
+\paragraph{Annotation and segmentation}\r
+Metadata also consist in temporal information such as \emph{time-coded makers} with comments and \emph{segmentation} according to ontology relevant for ethnomusicology (e.g. speech versus singing voice segment, chorus, ...)\r
+It should be notice that those annotations and segmentation can be done either by an human expert or by some audio processing automatic analysis (see Section~\ref{sec:Timeside}).\r
\r
+ \r
\section{TimeSide}\label{sec:Timeside}\r
- One specificity of the Telemeta architecture is to rely on an external component, \emph{TimeSide}, that offers audio player integration ``on the fly'' and advanced audio signal processing analysis capabilities.\r
++One specificity of the Telemeta architecture is to rely on an external component, \emph{TimeSide}, that offers audio player integration together with audio signal processing analysis capabilities.\r
++ \r
+Figure~\ref{fig:TimeSide_Archi} illustrates the overall architecture of \emph{TimeSide}.\r
\r
\begin{figure}[htbp]\r
\centering\r
- \includegraphics[width=12cm]{img/timeside_schema.pdf}\r
- \caption{TimeSide architecture}\r
+ \includegraphics[width=10cm]{img/timeside_schema.pdf}\r
+ \caption{TimeSide architecture}\label{fig:TimeSide_Archi}\r
\end{figure}\r
-\subsection{Audio management}\r
-Gstreamer, web player\r
-with enhance visualization (waveform, spectrogram)\r
-transcoding with smart streaming and caching methods (FLAC, OGG, MP3, WAV and WebM)\r
\r
-\subsection{Audio features extraction}\r
-Include reference audio feature tools : Aubio + Yaafe + Vamp\r
-\cite{yaafe_ISMIR2010,brossierPhD}\r
-flexible architecture \r
-\r
-\section{Current development and perspectives}\r
-interdisciplinarity is further enhance by the Music Information Retrieval, Speech technology \r
-Diadems project\r
-\subsection{Audio analysis}\r
-Development of tools to offer new audio analysis tool to ethnomusicologis research studies \r
-+ music similarity\r
-\r
-\subsection{Automatic segmentation and classification}\r
+\r
+\subsection{Audio management}\r
+TimeSide provides the following main features :\r
- \begin{itemize}\r
- \item \emph{Secure archiving, editing and publishing of audio files} over\r
- internet.\r
- \item Smart \emph{audio player} with enhance visualization (waveform, spectrogram)\r
- \item \emph{Multi-format support} : FLAC, OGG, MP3, WAV and more (through Gstreamer)\r
- % \item \emph{Playlist management} for all users with CSV data export\r
- \item "On the fly" \emph{audio analyzing, transcoding and metadata\r
+ \begin{itemize}\r
-\item singing / talking voice segment\r
-\item ...\r
++\item \emph{Secure archiving, editing and publishing of audio files} over\r
++ internet.\r
++\item Smart \emph{audio player} with enhance visualization (waveform, spectrogram)\r
++\item \emph{Multi-format support} : read all available audio and video formats through Gstreamer, transcoding with smart streaming and caching methods% (FLAC, OGG, MP3, WAV and WebM)\r
++ % \item \emph{Playlist management} for all users with CSV data export\r
++\item "On the fly" \emph{audio analyzing, transcoding and metadata\r
+ embedding} based on an easy plugin architecture\r
- \end{itemize}\r
+ \end{itemize}\r
\r
+\subsection{Audio features extraction}\r
+TimeSide incorporates some state-of-the-art audio feature extraction libraries such as Aubio, Yaafe and Vamp plugins \cite{brossierPhD,yaafe_ISMIR2010,vamp-plugins}.\r
+This feature extraction capability enable to automatically analyzes every sound items in a given collection and display the results as a support to ethnomusicological studies.\r
+Further works on that subject will incorporate advance Music Information Retrieval methods to provide automatic annotation and segmentation together with similarity analysis.\r
+\r
+\section{Conclusion - Purpose of the demonstration}\r
+The demonstration aims at presenting the features offered by \emph{Telemeta} as detailed in Section~\ref{sec:Telemeta} in the context of ethnomusicological sound archives \cite{telemetaCREM}. It focuses on the enhance and collaborative user-experience for accessing the audio items and associated metadata and on the possibility for the expert user to further enrich those metadata.\r
+Another goal of this demonstration is to present the integrated audio analysis tools described in Section~\ref{sec:Timeside}\r
+\r
\r
\subsubsection*{Acknowledgments.} \r
The authors would like to thanks all the people that have been involved in \emph{Telemeta} specification and development or have provide appreciated thoughts during discussions.\r
projects / telemeta-doc.git / commitdiff