A Simple Approach to Unsupervised Speaker Indexing

Ofoegbu, U.O.; Iyer, Ananth N.; Yantorno, Robert E.; Smolenski, Brett Y.

doi:10.1109/ispacs.2006.364901

Cited by 2 publications

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In previous research, which dealt with developing a speaker indexing system for telephone conversations, the effect on speaker differentiation of increasing the number of voiced phonemes used in forming speaker models was analyzed. Similar results were obtained, as the indexing accuracy increased with increasing number of voiced phonemes up to 5 (about 1 second of voiced speech), and then began to decrease (Ofoegbu et al 2006a). These observations imply that, for "blind" speaker comparisons (that is, when no information about speaker segmentation is known) with telephone conversations, a distance measure that performs well with little data is desirable.…”

Section: Discussionsupporting

confidence: 79%

“…In telephone conversations, homogeneous speaker utterances are short because speakers change relatively frequently (Iyer et al 2006;Ofoegbu et al 2006a); moreover, information about speaker change points is not usually available. Consequently, increasing the length of data segments used in comparison does not necessarily result in an increase in speaker differentiation performance, as data from more than one speaker could be combined in the same segment.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Speaker distinguishing distances: a comparative study

Iyer¹,

Ofoegbu²,

Yantorno

et al. 2007

Int J Speech Technol

Self Cite

View full text Add to dashboard Cite

Speaker discrimination is a vital aspect of speaker recognition applications such as speaker identification, verification, clustering, indexing and change-point detection. These tasks are usually performed using distance-based approaches to compare speaker models or features from homogeneous speaker segments in order to determine whether or not they belong to the same speaker. Several distance measures and features have been examined for all the different applications, however, no single distance or feature has been reported to perform optimally for all applications in all conditions. In this paper, a thorough analysis is made to determine the behavior of some frequently used distance measures, as well as features, in distinguishing speakers for different data lengths. Measures studied include the Mahalanobis distance, Kullback-Leibler (KL) distance, T 2 statistic, Hellinger distance, Bhattacharyya distance, Generalized Likelihood Ratio (GLR), Levenne distance, L 2 and L ∞ distances. The Mel-Scale Frequency Cepstral Coefficient (MFCC), Linear Predictive Cepstral Coefficients (LPCC), Line Spectral Pairs (LSP) and the Log Area Ratios (LAR) comprise the features investigated. The usefulness of these measures is studied for different data lengths. Generally, a larger data size for each speaker results in better speaker A.N. Iyer ( ) differentiating capability, as more information can be taken into account. However, in some applications such as segmentation of telephone data, speakers change frequently, making it impossible to obtain large speaker-consistent utterances (especially when speaker change-points are unknown). A metric is defined for determining the probability of speaker discrimination error obtainable for each distance measure using each feature set, and the effect of data size on this probability is observed. Furthermore, simple distancebased speaker identification and clustering systems are developed, and the performances of each distance and feature for various data sizes are evaluated on these systems in order to illustrate the importance of choosing the appropriate distance and feature for each application. Results show that for tasks which do not involve any limitation of data length, such as speaker identification, the Kullback Leibler distance with the MFCCs yield the highest speaker differentiation performance, which is comparable to results obtained using more complex state-of-the-art speaker identification systems. Results also indicate that the Hellinger and Bhattacharyya distances with the LSPs yield the best performance for small data sizes.

show abstract

Section: Discussionsupporting

confidence: 79%

Section: Discussionmentioning

confidence: 99%

Speaker distinguishing distances: a comparative study

Iyer¹,

Ofoegbu²,

Yantorno

et al. 2007

Int J Speech Technol

Self Cite

View full text Add to dashboard Cite

show abstract

Unsupervised Indexing of Conversations with Short Speaker Utterances

Ofoegbu

Iyer

Yantorno

et al. 2007

2007 IEEE Aerospace Conference

View full text Add to dashboard Cite

Two speaker indexing system for conversations are presented in this paper. The first method involves indexing two-speaker conversations. In this method, two reference models are judiciously chosen from the conversation such that they represent the two different speakers. Models are then matched to the reference speakers using distance-based comparisons. The second technique is based on first determining the number of participants in the conversation using a speaker count method termed the "Residual Ratio Algorithm" (RRA), and then indexing based on this count. The RRA involves an elimination process in which speech segments matching a chosen set of reference models are sequentially removed from the conversation and the relative amount of residual speech is observed to determine the count. The distance measures used in comparing models include the Bhattacharya distance, the T-Square statistics and the Mahalanobis distance. Speaker comparison decisions of all three distances are combined to improve the accuracy of the system. Linear Predictive Cepstral Coefficients of voiced phonemes are used in forming speaker models. The twospeaker indexing technique was able to yield an indexing accuracy of up to 95% when evaluated using SWITCHBOARD data. The counting-indexing technique resulted in a maximum indexing accuracy of about 91% when tested on artificial conversations generated from HTIMIT data. 12

show abstract

A Simple Approach to Unsupervised Speaker Indexing

Cited by 2 publications

References 9 publications

Speaker distinguishing distances: a comparative study

Speaker distinguishing distances: a comparative study

Unsupervised Indexing of Conversations with Short Speaker Utterances

Contact Info

Product

Resources

About