Dear All,
Sorry for the delay in circulating these. Please treat as a draft and
suggest amendments, additions or clarifications.
I was a little concerned as I wrote these up that the "actions" arising
seemed to be largely collective -- so I hope this doesn't mean nobody takes
responsibility!! What I've done is to highlight things which look like
actions, so please look carefully at these to see if you should be taking
the initiative. Hope the highlighting survives transform to rtf...
All the best,
Jill
--=====================_907252737==_
Content-Type: application/rtf; charset="us-ascii"
Content-Disposition: attachment; filename="ProminsEdinbrtf.rtf"
{\rtf1\ansi \deff4\deflang1033{\fonttbl{\f1\froman\fcharset2\fprq2 Symbol;}{\f4\froman\fcharset0\fprq2 Times New Roman;}}{\colortbl;\red0\green0\blue0;\red0\green0\blue255;\red0\green255\blue255;\red0\green255\blue0;
\red255\green0\blue255;\red255\green0\blue0;\red255\green255\blue0;\red255\green255\blue255;\red0\green0\blue128;\red0\green128\blue128;\red0\green128\blue0;\red128\green0\blue128;\red128\green0\blue0;\red128\green128\blue0;\red128\green128\blue128;
\red192\green192\blue192;}{\stylesheet{\widctlpar \f4\lang2057 \snext0 Normal;}{\*\cs10 \additive Default Paragraph Font;}{\s15\widctlpar\tqc\tx4153\tqr\tx8306 \f4\lang2057 \sbasedon0\snext15 header;}{\s16\widctlpar\tqc\tx4153\tqr\tx8306 \f4\lang2057
\sbasedon0\snext16 footer;}}{\info{\title ProSynth: Project meeting minutes }{\author J.House}{\operator J.House}{\creatim\yr1998\mo10\dy1\hr16\min26}{\revtim\yr1998\mo10\dy1\hr16\min38}{\version3}{\edmins3}{\nofpages2}{\nofwords654}{\nofchars3732}
{\*\company Phonetics,UCL}{\vern57443}}\paperw11906\paperh16838 \widowctrl\ftnbj\aenddoc\formshade \fet0\sectd \linex0\headery709\footery709\colsx709\endnhere {\header \pard\plain \s15\qr\widctlpar\tqc\tx4153\tqr\tx8306 \f4\lang2057 {\i
ProSynth, Edinburgh, 23/9/98}
\par }{\*\pnseclvl1\pnucrm\pnstart1\pnindent720\pnhang{\pntxta .}}{\*\pnseclvl2\pnucltr\pnstart1\pnindent720\pnhang{\pntxta .}}{\*\pnseclvl3\pndec\pnstart1\pnindent720\pnhang{\pntxta .}}{\*\pnseclvl4\pnlcltr\pnstart1\pnindent720\pnhang{\pntxta )}}{\*\pnseclvl5
\pndec\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl6\pnlcltr\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl7\pnlcrm\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl8\pnlcltr\pnstart1\pnindent720\pnhang
{\pntxtb (}{\pntxta )}}{\*\pnseclvl9\pnlcrm\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}\pard\plain \qc\widctlpar \f4\lang2057 {\b\fs32 ProSynth: Project meeting minutes
\par }\pard \widctlpar
\par Edinburgh, Heriot Watt University, preceding EPSRC workshop
\par
\par Date: 23 September 1998
\par
\par \pard \widctlpar Present: John, Richard, Paul (York), Sarah (Cambridge), Jill, Mark, Jana , Alex (UCL)
\par \pard \widctlpar Apologies: Sebastian
\par Chair: John, who welcomed Jana to the project.
\par Minuted by: Jill
\par
\par {\b 1. Current progress}
\par
\par {\pntext\pard\plain 1.1 \tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlbody\pndec\pnb0\pni0\pnf4\pnfs24\pnstart1\pnindent283\pnhang{\pntxtb 1.}{\pntxta }}York.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}John and Paul have produced paper and poster for EPSRC workshop.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}
Paul has been searching YorkTalk code for durational model to test against database. A range of relevant structural contexts identified. Will soon be able to extract and codify knowledge from YT classification trees, then search database for patterns.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}John has been computing durations and testing them; morphological issues (e.g. prefixes/suffixes) and f0 alignment need attent
ion. Soon be able to test in PSOLA.
\par \pard \widctlpar
\par {\pntext\pard\plain 1.2 \tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlbody\pndec\pnb0\pni0\pnf4\pnfs24\pnstart2\pnindent283\pnhang{\pntxtb 1.}{\pntxta }}UCL.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Jill has circulated current version of Intonational Phrase grammar, to sit on top of Richard\rquote
s Syllable grammar. It was felt that it would be better to retain most of the redundancy it contains, rather than to eliminate it. Use of the attributes {\i heavy}, {\i strong} and {\i headed} on feet should allow relevant structures to be identified.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Jana has made a start on labelling database files in ESPS, using scripts to make the system more friendly.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Mark presented a document repo
rting (i) Improvements to search tool and syllable labelling; (ii)Preliminary f0 modelling; (iii) MBROLA-compatible output; (iv) Prosodic manipulation. This requires 4 separate bits of data, so can currently be done in SFS but not ESPS.
\par \pard \widctlpar
\par {\pntext\pard\plain 1.3 \tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlbody\pndec\pnb0\pni0\pnf4\pnfs24\pnstart3\pnindent283\pnhang{\pntxtb 1.}{\pntxta }}Cambridge
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}
Paper on PROCSY by Heid & Hawkins circulated (ESCA workshop), and all current relevant ProSynth papers prepared for display at workshop [Alex\rquote s added].
\par {\pntext\pard\plain\f1 \'b7\tab}Continuing work on HLSyn copy synthesis to improve rules and make Ag parameter work properly. Current work includes:
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li1003\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Investigating what one can get in terms of spectral quality and edge phenomena by building in more realistic constriction areas.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li1003\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Looking at stressed vs unstressed syllables
\par {\pntext\pard\plain\f1 \'b7\tab}Studying articulatory trajectories in relation to vowel height/stress
\par {\pntext\pard\plain\f1 \'b7\tab}Exploring types of compression (timing; constriction areas)
\par {\pntext\pard\plain\f1 \'b7\tab}Modularising for testing hypotheses.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Perceptual tests to be in place by December.
\par \pard \widctlpar
\par {\pntext\pard\plain\b 2. \tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlbody\pndec\pnb1\pni0\pnf4\pnfs24\pnstart2\pnindent283\pnhang{\pntxta . }}{\b Issues arising
\par {\pntext\pard\plain\f1 \'b7\tab}}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Further meetings required.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li1003\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}{\highlight7 RAs need to get together} to discuss various issues: e.g. labelling criteria
, ESPS scripts, use of ESPS ~ SFS, integrating formal representation (database with macrostructural descriptions).
\par {\pntext\pard\plain\f1 \'b7\tab}{\highlight7 Longer meetings to be scheduled} to enable everyone to compare notes and progress. {\highlight7 PIs to attend meetings relevant}
to their issues; there are matters which Mark and Sebastian need to talk through.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li1003\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Cambridge is probably the most convenient venue; {\highlight7 first meeting to be arranged for mid-October.}
\par {\pntext\pard\plain\f1\highlight7 \'b7\tab}\pard \fi-283\li1003\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}{\highlight7 Participants to pre-circulate anything they want to bring to meeting.
\par {\pntext\pard\plain\f1 \'b7\tab}Conference phone calls to be used where these would be effective.}
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}{\highlight7 York and Cambridge to investigate use of SFS} to enable prosody manipulation to work.
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}{\highlight7 Workplan to be revised and adjusted over e-mail as soon as possible.}
\par \pard \widctlpar
\par {\pntext\pard\plain\b 3. \tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlbody\pndec\pnb1\pni0\pnf4\pnfs24\pnstart3\pnindent283\pnhang{\pntxta . }}{\b Publications and conferences
\par {\pntext\pard\plain\b\f1 \'b7\tab}}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}ICPhS: abstract deadlines imminent (15 October). {\highlight7 We should submit site-specific abstracts as we see fit.}{\b
\par {\pntext\pard\plain\b\f1 \'b7\tab}}{\highlight7 Need to plan big paper for refereed journal}. Timing is crucial because follow-up grant proposal due at latest Dec \rquote 99: we should aim to have paper {\i accepted} by then.{\b
\par {\pntext\pard\plain\b\f1 \'b7\tab}}\pard \fi-283\li1003\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Proposed schedule: start work in Feb, following perceptual test results, and submit in June. {\b
\par {\pntext\pard\plain\b\f1 \'b7\tab}}Possible journals: Speech Communication; Computer Speech & Language; JASA \endash decide which when the paper is taking shape. {\b
\par {\pntext\pard\plain\b\f1 \'b7\tab}}\pard \fi-283\li1003\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}
Content: to include account of how hierarchy integrates all the different components. Use ICSLP paper as outline; otherwise determine content by end Feb and verify against workplan.{\b
\par {\pntext\pard\plain\f1 \'b7\tab}}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}{\highlight7 Everyone should \ldblquote bounce\rdblquote papers/abstracts round the whole group} before publication.
\par \pard \widctlpar
\par {\pntext\pard\plain\b 4. \tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlbody\pndec\pnb1\pni0\pnf4\pnfs24\pnstart4\pnindent283\pnhang{\pntxta . }}{\b Ideas for project development
\par {\pntext\pard\plain\f1 \'b7\tab}}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Need to be planning ahead for grant extension (1999).
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li283\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}Need to identify possible \ldblquote satellite\rdblquote
projects to attract further grant applications. These need to be flagged up at workshop presentation and on our website. Possible projects might include:
\par {\pntext\pard\plain\f1 \'b7\tab}\pard \fi-283\li1003\widctlpar{\*\pn \pnlvlblt\pnf1\pnstart1\pnindent283\pnhang{\pntxtb \'b7}}implementing speaker-independent f0
\par {\pntext\pard\plain\f1 \'b7\tab}incorporating creak/voice quality effects
\par {\pntext\pard\plain\f1 \'b7\tab}work on articulation rate for English
\par {\pntext\pard\plain\f1 \'b7\tab}signal processing: work on prosody vocoders for presentation at COST
\par {\pntext\pard\plain\f1 \'b7\tab}doing concatenative synthesis using HLSyn
\par {\pntext\pard\plain\f1 \'b7\tab}influence of fine acoustic detail on lexical access
\par {\pntext\pard\plain\f1 \'b7\tab}effect of word boundaries within prosodic constituents
\par {\pntext\pard\plain\f1 \'b7\tab}function words and content words: two systems or one?
\par {\pntext\pard\plain\f1 \'b7\tab}Latinate vs Germanic timing.
\par \pard \widctlpar
\par
\par {\i\highlight7 Date of next meeting to be determined.}
\par }
--=====================_907252737==_
Content-Type: text/plain; charset="us-ascii"
--=====================_907252737==_--