506 INDEX
speech-to-speech translation (S2ST) (continued )
coupling speech recognition and translation,
341–347
lattice coupling, 343–347
overview, 341
removing disfluencies, 341–343
overview, 317–320
language portability, 319–320
overview, 317–318
speech translation strategies, 318–319
speech-to-text (STT), 48, 307–309
speech translation approaches, 320–340
comparing translation strategies, 337–340
EDTRL versus direct statistical translation,
339–340
Interlingua versus direct statistical
translation, 337–339
overview, 337
English, used as Interlingua, 334–337
connecting translation steps, 336–337
linguistically enhanced English as
Interlingua, 335–336
overview, 334
standardized and simplified English as
Interlingua, 335
Interlingua based speech translation,
320–324
interactively learning semantic
grammars, 322–324
Interlingua design, 321–322
overview, 320
semantic grammars, 322
overview, 320
statistical direct translation, 324–330
decoder, 327–330
overview, 324
phrase alignment models, 325–327
word alignment models, 324–325
statistical translation using formal
Interlingua, 330–334
decoding trees, 333–334
language model for trees, 331
overview, 330
translation models for trees, 331–333
speech translation strategies, 318–319,
394–395
SPEECON, 65–66, 70
SPICE, 91
spoken dialog systems (SDS), 399
spoken forms, versus written forms, 180
Spoken Language Database (SLDB), 372–374
spoken language-specific vocabulary items,
145–146
spoken response generation, ISIS system of, 413
SRE2004 program, 52
SRE2005 program, 52
SSMT (speech-to-speech machine translation
system), 318
SSS (successive state splitting) algorithm, 349
Standardized Character Encoding, 203
state-based dialog management, 404
Statistical ATR Translator (SAT), 360–361
statistical direct translation, 324–330
decoding approach, 327–330
overview, 324
phrase alignment models, 325–327
word alignment models, 324–325
statistical language modeling, 169–174
language models for new languages
and multilingual applications, 173–174
overview, 169–173
Statistical Machine Translation (SMT), 324,
339, 344, 351, 365
statistical translation using formal Interlingua,
330–334
decoding trees, 333–334
language model for trees, 331
overview, 330
translation models for trees, 331–333
story-specific language modeling, 196
Strange Corpus, 286
STT (speech-to-text), 48, 307–309
subdialog, 419
subject-verb-object (SVO ), 8
successive state splitting (SSS) algorithm, 349
support vector machines (SVMs), 249–251
SVMs (support vector machines), 249–251
SVO (subject-verb-object), 8
Switchboard corpus, 34
Switchboard-II, 47–48
syllabic systems, 25
Synonym Generator, 336
system-initiative dialogs, 400
T
tagging, 202
Talkbank corpora, 40

Get Multilingual Speech Processing now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.