Modified DISTBIC algorithm for sp
نویسنده
چکیده
The paper deals with the problem of automatic speaker change detection. A metric-based algorithm, called MDISTBIC, which means Modified DISTBIC, is proposed in this paper. The algorithm originates from the DISTBIC algorithm and modifies it in order to reach a higher efficiency. Both the DISTBIC and the MDISTBIC methods are tested in a number of experiments. As the results show, the MDISTBIC algorithm is more efficient than the DISTBIC algorithm in a majority of tests.
منابع مشابه
Modified DISTBIC algorithm for speaker change detection
The paper deals with the problem of automatic speaker change detection. A metric-based algorithm, called MDISTBIC, which means Modified DISTBIC, is proposed in this paper. The algorithm originates from the DISTBIC algorithm and modifies it in order to reach a higher efficiency. Both the DISTBIC and the MDISTBIC methods are tested in a number of experiments. As the results show, the MDISTBIC alg...
متن کاملImproving SVF with DISTBIC for Phoneme Segmentation
In this paper we examine an application for phoneme segmentation of DISTBIC, a two-pass, textindependent method traditionally used for speaker segmentation. The novelty of this paper is its experimentation with use of the spectral variation function (SVF), a simple non-parametric method for phone segmentation, as a replacement for the distance measure of the first pass of DISTBIC. In doing so w...
متن کاملConvergence and Stability of Modified Random SP-Iteration for A Generalized Asymptotically Quasi-Nonexpansive Mappings
The purpose of this paper is to study the convergence and the almost sure T-stability of the modied SP-type random iterative algorithm in a separable Banach spaces. The Bochner in-tegrability of andom xed points of this kind of random operators, the convergence and the almost sure T-stability for this kind of generalized asymptotically quasi-nonexpansive random mappings are obtained. Our result...
متن کاملWavelet-based speaker change detection in single channel speech data
Speaker segmentation is the task of finding speaker turns in an audio stream. We propose a metric-based algorithm based on Discrete Wavelet Transform (DWT) features. Principal component analysis (PCA) or linear discriminant analysis (LDA) [1] are further used to reduce the dimensionality of the feature space and remove redundant information. In the experiments our methods referred to as DWT-PCA...
متن کاملDISTBIC: A speaker-based segmentation for audio data indexing
In this paper, we address the problem of speaker-based segmentation, which is the ®rst necessary step for several indexing tasks. It aims to extract homogeneous segments containing the longest possible utterances produced by a single speaker. In our context, no assumption is made about prior knowledge of the speaker or speech signal characteristics (neither speaker model, nor speech model). How...
متن کامل