Publication

Robust HMM-Based Speech/Music Segmentation

Hervé Bourlard
2001
Reports

Abstract

In this paper we present a new approach towards high performance speech/music segmentation on realistic tasks related to the automatic transcription of broadcast news. In the approach presented here, the local probability density function (PDF) estimators trained on clean microphone speech are used as a channel model at the output of which the entropy and ``dynamism'' will be measured and integrated over time through a 2-state (speech and and non-speech) hidden Markov model (HMM) with minimum duration constraints. The parameters of the HMM are trained using the EM algorithm in a completely unsupervised manner. Different experiments, including a variety of speech and music styles, as well as different segment durations of speech and music signals (real data distribution, mostly speech, or mostly music), will illustrate the robustness of the approach, which in each case achieves a frame-level accuracy greater than 94%.

Official source

https://infoscience.epfl.ch/entities/publication/19218012-6979-4e60-926b-6fe5a8d2e2bc

About this result

This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.

Graph Chat

Chat with Graph Search

Ask any question about EPFL courses, lectures, exercises, research, news, etc. or try the example questions below.

DISCLAIMER: The Graph Chatbot is not programmed to provide explicit or categorical answers to your questions. Rather, it transforms your questions into API requests that are distributed across the various IT services officially administered by EPFL. Its purpose is solely to collect and recommend relevant references to content that you can explore to help you answer your questions.

Hervé Bourlard
2001
Reports

Abstract

Official source

https://infoscience.epfl.ch/entities/publication/19218012-6979-4e60-926b-6fe5a8d2e2bc

About this result

Related concepts (32)

Music

Music is generally defined as the art of arranging sound to create some combination of form, harmony, melody, rhythm, or otherwise expressive content. Definitions of music vary depending on culture, though it is an aspect of all human societies and a cultural universal. While scholars agree that music is defined by a few specific elements, there is no consensus on their precise definitions. The creation of music is commonly divided into musical composition, musical improvisation, and musical performance, though the topic itself extends into academic disciplines, criticism, philosophy, and psychology.

Music genre

A music genre is a conventional category that identifies some pieces of music as belonging to a shared tradition or set of conventions. It is to be distinguished from musical form and musical style, although in practice these terms are sometimes used interchangeably. Music can be divided into genres in varying ways, such as popular music and art music, or religious music and secular music. The artistic nature of music means that these classifications are often subjective and controversial, and some genres may overlap.

Arabic music

Arabic music (al-mūsīqā al-ʿarabīyyah) is the music of the Arab world with all its diverse music styles and genres. Arabic countries have many rich and varied styles of music and also many linguistic dialects, with each country and region having their own traditional music. Arabic music has a long history of interaction with many other regional musical styles and genres. It represents the music of all the peoples that make up the Arab world today.

Related publications (13)

Robust HMM-Based Speech/Music Segmentation

Chat with Graph Search

Towards Learning Emotion Information From Short Segments of Speech

Sequential Outlier Detection Based on Incremental Decision Trees

A Probabilistic Framework for Multiple Speaker Localization