[hts-users:02319] Re: Visual synthesis with HMM
- Subject: [hts-users:02319] Re: Visual synthesis with HMM
- From: Blaise Potard <bpotard@xxxxxxxxx>
- Date: Wed, 16 Dec 2009 11:31:59 +0000
- Delivered-to: hts-users@xxxxxxxxxxxxxxx
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:in-reply-to:references :date:message-id:subject:from:to:content-type :content-transfer-encoding; bh=nOP8F++e+VUTJgssYWmgNu8p1JQTrBE3yokziPuXtEE=; b=nElLaHGggbt6myGJ5bVoJAOXUHD4Y21qmz7qe7k9T63w07Hw63krcZ8tIHpRqEd9Xv G/sVSnIWedvvdrcy2yJucIucDC/vcm04U+dCHvAZawnUcrHpSfSIo72oqr4+Qi6ngFqC LveuBrDc9vZPDDN2zECTARB06mXI7YsIvMOjw=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type:content-transfer-encoding; b=nRrXyVfpmk7KVcRhYsXsBTrlvYdDKoa0Ky2/cCMm9w056QeQNOMWwa5o3fRnTh0ahh RW40o0iMxESZcf6HHXRMGdR27ARHci2mP+zXxjoLwiaS86x+utj3OxNJBKn4MQIvW23M 6j0D6hVHpYrTkMZpOlSoIOLhMH9M9c0ohAFxM=
Hello,
Yes, this is a small bug in the HTSdemo script, which fails when there
is only one stream. You can correct it simply by changing two lines:
< if ($set eq "dur") {
---
> if ($#{$ref{$set}} eq 0 ) {
It occurs twice if I remember correctly.
in make_edfile_upmix
and make_edfile_untie
(the main idea is that when there is only one stream, the macros
become more generic; to handle this particular case we just need to do
the same as for the duration, which has always only one stream).
Hope that helps!
Blaise
P.S.: you should probably consider reducing the number of parameters
you use (through PCA or so), 68 might be a bit too much.
2009/12/16 Borja Martinez Piazuelo <borjam@xxxxxxxxxxxxx>:
> Hi,
>
> I am trying to make visual synthesis with HHM by generating lip-sync with
> the HTS-demo_CMU-ARCTIC-SLT.
> I’m using FAPs (facial animation parameters). For each frame I have 68
> parameters for the position of each point of the face instead of mgc and
> f0, to animate a 3D.
>
> When it comes to the CXCL2 function I get the next error:
>
> // construct decision trees
> TB 0.00 fap_s2_ {*.state[2].stream[1-1]}
> Tree based clustering based on MDL criterion, threshold=7.963560e+02
> ERROR [+7036] NewMacro: macro or model name fap_s2_4 already exists
> FATAL ERROR - Terminating program /home/Borja/tools/htk/HTKTools/HHEd
> Error in /home/Borja/tools/htk/HTKTools/HHEd -A -C
> /home/Borja/tools/voices/HTS-demo_CMU-ARCTIC-SLT_Borja/configs/trn.cnf -D
> -T 1 -p -i -H
> /home/Borja/tools/voices/HTS-demo_CMU-ARCTIC-SLT_Borja/models/qst001/ver1/cmp/re_clustered.mmf
> -m -a 1.0 -w
> /home/Borja/tools/voices/HTS-demo_CMUARCTICSLT_Borja/models/qst001/ver1/cmp/re_clustered.mmf
> /home/Borja/tool
> s/voices/HTS-demo_CMU-ARCTIC-SLT_Borja/edfiles/qst001/ver1/cmp/cxc_fap.hed.untied
> /home/Borja/tools/voices/HTS-demo_CMU-ARCTIC-SLT_Borja/data/lists_fap/full.list
>
> I don’t know if it’s due to the fact that I’m using only one stream of
> parameters (FAPs) instead of two (mgc and f0). Can someone help me or tell
> the possibles causes of this error?
>
> Thanks!
>
> Borja
- Follow-Ups
-
- [hts-users:02320] Re: Visual synthesis with HMM, Borja Martinez Piazuelo
- References
-
- [hts-users:02316] Visual synthesis with HMM, Borja Martinez Piazuelo