CORPUS-BASED SPEECH SYNTHESIS BASED ON SEGMENT RECOMBINATION

Patent №

US 7,567,896

Granted

2009-07-28

Filed 2005

Owner

SCANSOFT, INC.

Lab

—

AI components

ml · nlp · speech

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

11037545

A system and method generate synthesized speech through concatenation of speech segments that are derived from a large prosodically-rich corpus of speech segments including using an additional dictionary of speech segment identifier sequences.

Machine learning Natural language SpeechG10L 13/06G10L 13/07

Google Patents

AI classification

Natural language1.00

Speech1.00

Machine learning0.97

Knowledge representation0.08

AI hardware0.01

Evolutionary computation0.00

Planning0.00

Vision0.00

Ownership

SCANSOFT, INC.

assignment · 159490211

Assignors

COORMAN, GEERT, POLLET, VINCENT, VAN GERVEN, STEFAAN, DE BOCK, MARIO, VAN COILE, BERT, DE MOORTEL, JAN

On an employer assignment, the assignors are typically the inventors.

From the same owner

RANDOM CONFIRMATION IN SPEECH BASED SYSTEMSUS 7,844,465 · 2010 RANDOM CONFIRMATION IN SPEECH BASED SYSTEMSUS 7,788,103 · 2010 DISCRIMINATIVE TRAINING OF HIDDEN MARKOV MODELS FOR CONTINUOUS SPEECH RECOGNITIONUS 7,672,847 · 2010 METHOD AND APPARATUS FOR EFFICIENT SEGMENTATION OF COMPOUND WORDS USING PROBABILISTIC BREAKPOINT TRAVERSALUS 7,610,189 · 2009 LANGUAGE MODEL FOR USE IN SPEECH RECOGNITIONUS 7,584,102 · 2009 AUTOMATIC INSERTION OF NON-VERBALIZED PUNCTUATIONUS 7,580,838 · 2009 SYSTEMS AND METHODS FOR EVALUATING SPEAKER SUITABILITY FOR AUTOMATIC SPEECH RECOGNITION AIDED TRANSCRIPTIONUS 7,292,975 · 2007 SYSTEM AND METHOD FOR AUTOMATICALLY PROCESSING A USER'S REQUEST BY AN AUTOMATED ASSISTANTUS 7,242,758 · 2007

Cites (15)

SPEECH SYNTHESIS USING CONCATENATION OF SPEECH WAVEFORMSUS 7,219,060 · 2007 SYSTEM AND METHOD OF PROVIDING CONVERSATIONAL VISUAL PROSODY FOR TALKING HEADSUS 7,136,818 · 2006 CORPUS-BASED PROSODY TRANSLATION SYSTEMUS 7,069,216 · 2006 SYNTHESIS UNIT SELECTION APPARATUS AND METHOD, AND STORAGE MEDIUMUS 6,980,955 · 2005 SPEECH SYNTHESIS USING CONCATENATION OF SPEECH WAVEFORMSUS 6,665,641 · 2003 SPEECH SYNTHESISUS 5,978,764 · 1999 METHOD AND SYSTEM FOR SYNTHESIZING SPEECHUS 5,970,453 · 1999 COMMUNICATION SYSTEM AND METHOD USING A SPEAKER DEPENDENT TIME-SCALING TECHNIQUEUS 5,920,840 · 1999 METHOD AND SYSTEM OF RUNTIME ACOUSTIC UNIT SELECTION FOR SPEECH SYNTHESISUS 5,913,193 · 1999 TEXT TO SPEECH SYSTEMUS 5,774,854 · 1998 METHOD AND SYSTEM FOR TIME SCALE MODIFICATION UTILIZING FEATURE VECTORS ABOUT ZERO CROSSING POINTSUS 5,749,064 · 1998 AUTOMATED VOICE SYNTHESIS EMPLOYING ENHANCED PROSODIC TREATMENT OF TEXT, SPELLING OF TEXT AND RATE OF ANNUNCIATIONUS 5,652,828 · 1997 WAVEFORM BLENDING TECHNIQUE FOR TEXT-TO-SPEECH SYSTEMUS 5,490,234 · 1996 METHOD AND APPARATUS FOR SPEECH SYNTHESIS BASED ON PROSODIC ANALYSISUS 5,384,893 · 1995 GENERATING SPEECH FROM DIGITALLY STORED COARTICULATED SPEECH SEGMENTSUS 5,153,913 · 1992

Cited by (43)

PROBABILISTIC LANGUAGE MODELS FOR IDENTIFYING SEQUENTIAL READING ORDER OF DISCONTINUOUS TEXT SEGMENTSUS 11,769,111 · 2023 TEXT-TO-SPEECH FROM MEDIA CONTENT ITEM SNIPPETSUS 11,710,474 · 2023 VOICE-BASED AUTHENTICATIONUS 11,170,787 · 2021 TEXT-TO-SPEECH FROM MEDIA CONTENT ITEM SNIPPETSUS 11,114,085 · 2021 Speech Synthesis Using One or More Recurrent Neural Networks

Related research

CONCATENTIVE SPEECH SYNTHESIS USING SUB-PHONEME SEGMENTSOpenAlex Towards Optimal TTS CorporaSemantic Scholar Pronunciation Dictionary-Free Multilingual Speech Synthesis by Combining Unsupervised and Supervised Phonetic RepresentationsOpenAlex Speech Recognition with Augmented Synthesized SpeecharXiv Personalized End-to-End Mandarin Speech Synthesis using Small-sized CorpusOpenAlex

Full text

Google Patents