Synthesizing speech by converting phonemes to digital waveforms

• US 5,970,454 A
• Filed: 04/23/1997
• Issued: 10/19/1999
• Est. Priority Date: 12/16/1993
• Status: Expired due to Term

• Current Assignee: British Telecommunications PLC (BT Group PLC)
• Original Assignee: British Telecommunications PLC (BT Group PLC)
• Inventors: Breen, Andrew Paul
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): Opsasnick, Michael N.
• Application Number:
  US08/844,859
• Time in Patent Office: 909 Days
• Field of Search: 704/258, 704/260, 704/269, 704/231, 707/100
• US Class Current: 704/269
• CPC Class Codes: G10L 13/07 Concatenation rules

Abstract

Synthetic speech is generated by production of a digital waveform from a text in phonemes. A linked database is used which comprises an extended text in phonemes and its equivalent in the form of a digital waveform. The two portions of the database are linked by a parameter which establishes equivalent points in both the phoneme text and the digital waveform. The input text (in phonemes) is analyzed to locate a matching portion in the phoneme portion of the database. This matching utilizes exact equivalence of phonemes where this is possible; otherwise relation between phonemes is utilized. The selection process identifies input phonemes in context whereby improved conversions are obtained. Having analyzed the input exit into matching strings in the input form of the database beginning and ending parameters for the sections are established. The output text is produced by abutting sections of the digital waveform and defined by the beginning and ending parameters.

Claims

1. A method of converting an input signal representing a text in phonemes into an output digital waveform signal convertible into an acoustic synthesized speech waveform corresponding to said text, wherein said method makes use of a two-part database having an access section based on strings of phonemes and an output section containing digital waveforms corresponding to the linked access sections, wherein said method comprises:

matching a segment of said input signal to select the best match of strings contained in the access section said best match including an exact match for at least one internal phoneme, and

discarding at least the first and last phonemes of said best match to identify a shorter string of phonemes which is an exact match for a portion of said input signal.

2. A method as in claim 1 which includes:

forming a best match for a window of five phonemes of said input signal, and

discarding at least the first and last phonemes of said best match to identify an exact match for a string of one, two or three phonemes.

3. A method as in claim 2 wherein said forming a best match step includes an exact match for the third phoneme in the window of five phonemes.