A coupled HMM for solving the permutation problem.pdf (136.08 kB)
A coupled HMM for solving the permutation problem in frequency domain BSS
conference contribution
posted on 2010-02-04, 17:20 authored by Saeid Sanei, Wenwu Wang, Jonathon ChambersPermutation of the outputs at different frequency bins
remains as a major problem in the convolutive blind source
separation (BSS). In this work a coupled Hidden Markov
model (CHMM) effectively exploits the psychoacoustic
characteristics of signals to mitigate such permutation. A
joint diagonalization algorithm for convolutive BSS, which
incorporates a non-unitary penalty term within the crosspower
spectrum-based cost function in the frequency
domain, has been used. The proposed CHMM system
couples a number of conventional HMMs, equivalent to the
number of outputs, by making state transitions in each
model dependent not only on its own previous state, but
also on some aspects of the state of the other models. Using
this method the permutation effect has been substantially
reduced, and demonstrated using a number of simulation
studies.
History
School
- Mechanical, Electrical and Manufacturing Engineering
Citation
SANEI, S., WANG, W. and CHAMBERS, J.A., 2004. A coupled HMM for solving the permutation problem in frequency domain BSS. IN: Proceedings of 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP '04), Montreal, Quebec, 17-21 May, Vol. 5, pp. 565-8.Publisher
© IEEEVersion
- VoR (Version of Record)
Publication date
2004Notes
This is a conference paper [© IEEE]. It is also available from: http://ieeexplore.ieee.org/. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.ISBN
0780384849Language
- en