LINGUIST List 7.1434

Mon Oct 14 1996

Jobs: Positions in multi-language speech synthesis by rule

Editor for this issue: Ljuba Veselinova <lveselinemunix.emich.edu>


Directory

  1. Sue Hertz, Jobs in multi-language speech synthesis by rule

Message 1: Jobs in multi-language speech synthesis by rule

Date: Sat, 12 Oct 1996 10:27:29 EDT
From: Sue Hertz <hertzFIRE.DMLL.CORNELL.EDU>
Subject: Jobs in multi-language speech synthesis by rule
 POSITIONS AVAILABLE IN MULTI-LANGUAGE AND MULTI-DIALECT 
 SPEECH SYNTHESIS BY RULE

Eloquent Technology, Inc. has several positions available in
multi-language and multi-dialect text-to-speech synthesis. We are
looking for candidates at all levels of experience, with a background
in one or more of the following areas:

 - speech synthesis by rule (text-to-phoneme conversion or 
 phoneme-to-speech conversion)
 - speech analysis
 - acoustic phonetics
 - computational linguistics

A background in computer programming is desirable, but not
essential. Knowledge of German, French, Spanish, Italian,
Mandarin Chinese or British English will be especially useful. A
bachelor's degree is essential, and higher degrees are required
for some positions.

Eloquent Technology, Inc. (ETI) is a small text-to-speech company
in Ithaca, New York, the home of Cornell University. Ithaca is
located in the midst of the beautiful Fingerlakes region. ETI
presently consists of ten people, including a number of linguists
and computer scientists, and is expected to grow to fifteen
people over the next six months. ETI's main product is
ETI-Eloquence, a text-to-speech program currently available for
General American English. Other English dialects and languages
are being developed or planned. A demo version of ETI-Eloquence
that runs on PCs (Windows) is available upon request.

ETI uses a linguistically-oriented, rule-based approach to
text-to-speech synthesis. The ETI-Eloquence rule programs are
developed with the Delta System, a sophisticated software tool
consisting of a special programming language and interactive
environment designed for expressing phonological and phonetic
rules that operate on multi- tiered utterance representations
called "deltas". A delta consists of user-defined "streams"
(hence the name "delta") of coordinated units, such as
intonational phrases, syllables, words, phonemes, acoustic
parameter values, and durations. From the final delta for an
utterance, parameter values are generated for a formant-based
(Klatt-type) speech synthesizer. Delta provides sophisticated
formalisms for pattern-matching against multi-stream deltas and
manipulating them. Acoustic values can be modified
interactively, and the effects of the modifications listened to.

Further information about ETI, the models underlying the
ETI-Eloquence rule programs, and a general description of ETI's
approach to synthesis is available in the company white paper,
which we will send to you upon request.

See also ETI's web page at http://www.eloq.com

If you are interested, please send a detailed resume and cover
letter by e-mail (plain text is preferred) to Sue Hertz at:
 
 hertzeloq.com

or send fax copies to:

 607-266-7030

Alternatively, you can send a paper copy to

			Sue Hertz
			Eloquent Technology, Inc.
			2389 N. Triphammer Rd.
			Ithaca, New York 14850

Preference will be given to candidates who currently have the
ability to work in the United States and could start work soon.

ETI will consider applicants until all positions are filled.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue