[Subject Prev][Subject Next][Thread Prev][Thread Next][Date Index][Thread Index]

[hts-users:00664] Re: Formant emphasis for mel-generalized cepstrum


Thanks a lot Prof. Tokuda! From what I could glean from those papers,
I will need to implement a perceptual weighting filter and a
short-term postfilter that are applied to the MGC-LSP coefficients
before they enter MGLSA analysis. Is this correct?

                                                          Alex.

On 4/23/07, Keiichi Tokuda <tokuda@xxxxxxxxxxxx> wrote:
From: "Alexander Gutkin" <alexander.gutkin@xxxxxxxxx>
Subject: [hts-users:00659] Formant emphasis for mel-generalized cepstrum
Date: Mon, 23 Apr 2007 13:26:00 +0100
Message-ID: <6c7b80c30704230526g13b3db71x8aef373416efe985@xxxxxxxxxxxxxx>

> Does anyone know how to perform formant emphasis (postfiltering) for
> the cases when the input consists of coefficients generated using
> mel-generalized or mel-generalized LSP analysis? Essentially, I need
> an MGC or MGC-LSP equivalent of the current postfiltering() routine
> from Training.pl script.

See those papers:

  Kazuhito Koishida, Goh Hirabayashi, Keiichi Tokuda, and Takao
  Kobayashi, ``A 16kbit/s wideband CELP coder using
  mel-generalized cepstral analysis and its subjective
  evaluation,'' Proc. of International Conference on Spoken
  Language Processing (ICLSP-98), vol.6., pp.2583-2586, Sydney,
  Australia, Nov.-Dec., 1998.

  Kazuhito Koishida, Goh Hirabayashi, Keiichi Tokuda, and Takao
  Kobayashi, ``A wideband CELP speech coder at 16 kbit/s based
  on mel-generalized cepstral analysis,'' Proceedings of IEEE
  International Conference on Acoustics, Speech, and Signal
  Processing, vol.1, pp.161-164, Seattle, USA, May 1998.

Keiichi Tokuda
tokuda@xxxxxxxxxxxx
http://www.sp.nitech.ac.jp/~tokuda/


Follow-Ups
[hts-users:00669] Re: Formant emphasis for mel-generalized cepstrum, Keiichi Tokuda
References
[hts-users:00659] Formant emphasis for mel-generalized cepstrum, Alexander Gutkin
[hts-users:00660] Re: Formant emphasis for mel-generalized cepstrum, Keiichi Tokuda