David Jay Reich

US Patent:

6832196, Dec 14, 2004

Filed:

Mar 30, 2001

Appl. No.:

09/822657

Inventors:

David E. Reich - Jupiter FL

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G10L 2106

US Classification:

7042701, 704275, 704251, 704260

Abstract:

A method of dynamically formatting a speech menu construct can include a series of steps. A markup language document containing a reference to a server-side program can be provided. The server-side program can be programmed to dynamically format data using a voice-enabled markup language. A database can be accessed using the server-side program. The database can have a plurality of data items. Using the voice-enabled markup language, the selected data items can be formatted thereby creating speech menu items. The speech menu items can specify a speech menu construct resulting in a menu interface that is dynamically generated from data in data store, rather than being written by a programmer, and allows the user to âspeak to the data. â.

US Patent:

6839667, Jan 4, 2005

Filed:

May 16, 2001

Appl. No.:

09/858399

Inventors:

David E. Reich - Jupiter FL, US

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G10L 1522

US Classification:

704240, 704239, 704255

Abstract:

A method for performing speech recognition can include receiving user speech and determining a plurality of potential candidates. Each of the candidates can provide a textual interpretation of the speech. Confidence scores can be calculated for the candidates. The confidence scores can be compared to a predetermined threshold. Also, selected ones of the plurality of candidates can be presented to the user as alternative interpretations of the speech when none of the confidence scores is greater than the predetermined threshold. The selected ones of the plurality of candidates can have confidence scores above a predetermined minimum threshold, and thus can have confidence scores within a predetermined range.

US Patent:

7099828, Aug 29, 2006

Filed:

Nov 7, 2001

Appl. No.:

10/007615

Inventors:

Jeffrey S. Kobal - Lake Worth FL, US
Bruce D. Lucas - Mohecan Lake NY, US
David E. Reich - Jupiter FL, US

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G10L 21/00
G10L 13/00
G06F 3/00

US Classification:

704270, 704258, 715727

Abstract:

A method of generating pronunciation information can include graphically presenting at least one activatable visual identifier corresponding to individual ones of a plurality of phonemes. Responsive to a selection of one of the visual identifiers, pronunciation information can be generated in accordance with the selected visual identifier. The pronunciation information can be compiled responsive to a selection of one of the plurality of visual identifiers.

US Patent:

7308407, Dec 11, 2007

Filed:

Mar 3, 2003

Appl. No.:

10/378338

Inventors:

David E. Reich - Jupiter FL, US

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G10L 19/14

US Classification:

704266

Abstract:

A method for generating synthetic speech can include identifying a recording of conversational speech and creating a transcription of the conversational speech. Using the transcription, rather than a predefined script, the recording can be analyzed and acoustic units extracted. Each acoustic unit can include a phoneme and/or a sub-phoneme. The acoustic units can be stored so that a concatenative text-to-speech engine can later splice the acoustic units together to produce synthetic speech.

US Patent:

7640159, Dec 29, 2009

Filed:

Jul 22, 2004

Appl. No.:

10/896426

Inventors:

David E. Reich - Jupiter FL, US

Assignee:

Nuance Communications, Inc. - Burlington MA

International Classification:

G10L 15/02
G10L 15/04
G10L 15/28
G06F 17/20
G06F 17/27
G10L 17/00
H04M 1/64

US Classification:

704254, 704 8, 704 9, 704250, 704255, 379 8805, 379 8802, 379 8806

Abstract:

An accent compensative speech recognition system and related methods for use with a signal processor generating one or more feature vectors based upon a voice-induced electrical signal are provided. The system includes a first-language acoustic module that determines a first-language phoneme sequence based upon one or more feature vectors, and a second-language lexicon module that determines a second-language speech segment based upon the first-language phoneme sequence. A method aspect includes the steps of generating a first-language phoneme sequence from at least one feature vector based upon a first-language phoneme model, and determining a second-language speech segment from the first-language phoneme sequence based upon a second-language lexicon model.

US Patent:

7860491, Dec 28, 2010

Filed:

Dec 11, 2006

Appl. No.:

11/609160

Inventors:

David E. Reich - Jupiter FL, US
Fernando G. Concha - Coral Gables FL, US

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

H04M 3/00

US Classification:

455418, 455 412, 455 413, 4554141, 4554121, 455421, 4554321, 370338, 370216, 370229

Abstract:

A method of providing data content obtained from one or more network-connected nodes to a mobile computing device is provided. The method includes determining whether a previously-established network connection between the mobile computing device and a network-connected node has been interrupted after the mobile computing device receives a first set of data content and an identifier corresponding to a second set of data content from the network-connected node via the previously-established network connection. The method also includes establishing a wireless connection between the mobile computing device and another computing device if the previously-established network connection has been interrupted and the second set of data content is stored by the other computing device. The method further includes conveying the second set of data content to the mobile computing device from the other computing device via the wireless connection if the previously-established network connection has been interrupted and the second set of data content is stored at the other computing device.

US Patent:

8036893, Oct 11, 2011

Filed:

Jul 22, 2004

Appl. No.:

10/896405

Inventors:

David E. Reich - Jupiter FL, US

Assignee:

Nuance Communications, Inc. - Burlington MA

International Classification:

G10L 15/00
G10L 15/18
G06F 17/27
H04M 1/64

US Classification:

704257, 704 9, 704236, 379 8805

Abstract:

A system for use in speech recognition includes an acoustic module accessing a plurality of distinct-language acoustic models, each based upon a different language; a lexicon module accessing at least one lexicon model; and a speech recognition output module. The speech recognition output module generates a first speech recognition output using a first model combination that combines one of the plurality of distinct-language acoustic models with the at least one lexicon model. In response to a threshold determination, the speech recognition output module generates a second speech recognition output using a second model combination that combines a different one of the plurality of distinct-language acoustic models with the at least one distinct-language lexicon model.

US Patent:

8285546, Oct 9, 2012

Filed:

Sep 9, 2011

Appl. No.:

13/228879

Inventors:

David E. Reich - Jupiter FL, US

Assignee:

Nuance Communications, Inc. - Burlington MA

International Classification:

G10L 15/14
G10L 15/00
G10L 15/18
G10L 15/10
G10L 15/28
G10L 17/00

US Classification:

704257, 704250, 704254, 704255, 379 8808

Abstract:

A system for use in speech recognition includes an acoustic module accessing a plurality of distinct-language acoustic models, each based upon a different language; a lexicon module accessing at least one lexicon model; and a speech recognition output module. The speech recognition output module generates a first speech recognition output using a first model combination that combines one of the plurality of distinct-language acoustic models with the at least one lexicon model. In response to a threshold determination, the speech recognition output module generates a second speech recognition output using a second model combination that combines a different one of the plurality of distinct-language acoustic models with the at least one distinct-language lexicon model.