The Annals of Applied Statistics
- Ann. Appl. Stat.
- Volume 5, Number 2A (2011), 1020-1056.
A sticky HDP-HMM with application to speaker diarization
We consider the problem of speaker diarization, the problem of segmenting an audio recording of a meeting into temporal segments corresponding to individual speakers. The problem is rendered particularly difficult by the fact that we are not allowed to assume knowledge of the number of people participating in the meeting. To address this problem, we take a Bayesian nonparametric approach to speaker diarization that builds on the hierarchical Dirichlet process hidden Markov model (HDP-HMM) of Teh et al. [J. Amer. Statist. Assoc. 101 (2006) 1566–1581]. Although the basic HDP-HMM tends to over-segment the audio data—creating redundant states and rapidly switching among them—we describe an augmented HDP-HMM that provides effective control over the switching rate. We also show that this augmentation makes it possible to treat emission distributions nonparametrically. To scale the resulting architecture to realistic diarization problems, we develop a sampling algorithm that employs a truncated approximation of the Dirichlet process to jointly resample the full state sequence, greatly improving mixing rates. Working with a benchmark NIST data set, we show that our Bayesian nonparametric architecture yields state-of-the-art speaker diarization results.
Ann. Appl. Stat., Volume 5, Number 2A (2011), 1020-1056.
First available in Project Euclid: 13 July 2011
Permanent link to this document
Digital Object Identifier
Mathematical Reviews number (MathSciNet)
Zentralblatt MATH identifier
Fox, Emily B.; Sudderth, Erik B.; Jordan, Michael I.; Willsky, Alan S. A sticky HDP-HMM with application to speaker diarization. Ann. Appl. Stat. 5 (2011), no. 2A, 1020--1056. doi:10.1214/10-AOAS395. https://projecteuclid.org/euclid.aoas/1310562215
- Supplementary material: Notational conventions, Chinese restaurant franchises and derivations of Gibbs samplers. We present detailed derivations of the conditional distributions used for both the direct assignment and blocked Gibbs samplers, as well as the associated pseudo-code. The description of these derivations relies on the Chinese restaurant analogies associated with the HDP and sticky HDP-HMM, which are expounded upon in this supplementary material. We also provide a list of notational conventions used throughout the paper.