International Business Machines Corporation

Nuance Communications, Inc.

Stephen Graham Lawrence

Brian H Middleton

Michael Alan Picheny

John Brian Pickering

Maria Elena Smith

Raimo Bakis

Abraham P Ittycheriah

Nicholas David Butler

Ronald John Bowater

Edward Adam Epstein

Barry Neil Dow

David Andrew Clarke

Charles T Rutherfoord

Steven J Friedland

Hari Chittaluru

Michael A McMorran

A distributed voice recognition system () and method employs principles of bottom-up (i.e., raw input) and top-down (i.e., prediction based on past experience) processing to perform client-side and server-side processing by (i) at the client-side, replacing application data by a phonotactic table (); (ii) at the server-side, tracking separate confidence scores for matches against an acoustic model and comparison to a grammar; and (iii) at the server-side using a contention resolver () to weight the client-side and server-side results to establish a single output which represents the collaboration between client-side processing and server-side processing.

Distributed voice recognition system and method

In a text-to-speech system, a method of converting text-to-speech can include receiving a text input and comparing the received text input to at least one entry in a text-to-speech cache memory. Each entry in the text-to-speech cache memory can specify a corresponding spoken output. If the text input matches one of the entries in the text-to-speech cache memory, the cached speech output specified by the matching entry can be provided.

Method and system for text-to-speech caching

The present invention relates to establishing a communication channel between two communication systems having computer telephony integration (CTI). Many CTI systems are configured into incoming and outgoing lines according to the anticipated demands. Consequently, if sufficient outgoing or incoming capacity is unavailable at a given time to support a communication channel of required characteristics, the users of such a system must conventionally wait until sufficient capacity becomes available. However, the present invention determines who the intended addressee is and can instruct the CTI system associated with that addressee to instigate the establishment of a communication channel to the user who originally desired the connection. The instructions can be sent to the intended addressee using for example, another communication network, such as the Internet or other data communications network. This arrangement also allows the call to be placed in the reverse direction, where this is favourable for tariff reasons. The invention also includes the automatic selection of a telephone network carrier based on tariff data requested over a data network such as the Internet.

Communication method and system

A display terminal of the type which is dependent on a remote processor for its operation and for its main data storage, comprising an interactive language interpreter (24) (e.g. for BASIC), terminal processor, local storage, display and a switch for switching between a session (90) associated with the remote processor and a session (91) dedicated to servicing the interactive language interpreter (24). During an interactive language session screens of data associated with the host session may be displayed and then saved which enables host data to be processed using user programs in the interactive language.

Display terminal

A visual display unit with character overstrike capability includes a text store in which text is stored in data stream form. A formatter formats the stored text and loads character codes into a refresh buffer which together with a character generator display the characters on a raster scanned cathode ray tube. Control logic responsive to a backspace character combines the bit patterns corresponding to the characters preceding and succeeding the backspace character in a temporary store and then stores the composite bit pattern in the character generator and a pointer thereto in the refresh buffer. The control logic can be inhibited to allow the individual characters constituting the composite character to be displayed individually for editing purposes.

Visual display unit with character overstrike

A text terminal in which a text stream received on a communication link is segmented by a segmenter into chained variable-length segments which are stored in a random access text store. Each segment can contain at most one extended formatting command. Segments have slack areas to allow the subsequent insertion of text. A text editor receives keystroke signals from a keyboard and modifies the stored text stream, new segments being created as necessary for extended formatting commands or when slack areas are full. An interpreter/formatter is invoked when a received text stream has been segmented or when the stored text stream has been modified to format and map the stored text stream into a refresh buffer driving a cathode ray tube (CRT) display. The interpreter/formatter maintains a status index with condensed versions of the extended formatting commands stored in the text stream. Formatting is inhibited by the editor if it receives a keystroke requiring modification of the stored text. After editing segmented text is desegmented by a desegmenter and returned to the host along a communication link in the form in which it was received. Preferably, the terminal is implemented with a microprocessor which can also control standard data entry applications as well as text processing.

Text processing apparatus with editing of stored document at each

Growing community of inventors

Southampton, United Kingdom

Stephen Graham Lawrence