Abstract: The perceived quality of an audio signals obtained from very low bit-rate audio coding system is improved by using expanding quantizers and arithmetic coding in a transmitter and using complementary compression and arithmetic decoding in a receiver. An expanding quantizer is used to control the number of signal components that are quantized to zero and arithmetic coding is used to efficiently code the quantized-to-zero coefficients. This allows a wider bandwidth and more accurately quantized baseband signal to be conveyed to the receiver, which regenerates an output signal by synthesizing the missing components.

Abstract: Synthesizing speech sounds to express a short message received from a caller in a handset coupled to a hands free kit, includes handset circuitry for transferring an alarm signal to the hands free kit to generate an alarm to inform the user of the receipt of the short message, and for transferring the short message to the hands free kit when receiving a short message calling signal from the hands free kit. The short message calling signal is generated upon the detection of a predetermined voice command by the user indicating a desire to hear the short message. The hands free kit includes circuitry for synthesizing the speech sounds according to the short message received from the handset.

Abstract: A computer-implemented method is described for optimizing prompts for a speech-enabled application. The speech-enabled application is operable to receive communications from a number of users and communicate one or more prompts to each user to illicit a response from the user that indicates the purpose of the user's communication. The method includes determining a number of prompt alternatives (each including one or more prompts) to evaluate and determining an evaluation period for each prompt alternative. The method also includes automatically presenting each prompt alternative to users during the associated evaluation period and automatically recording the results of user responses to each prompt alternative. Furthermore, the method includes automatically analyzing the recorded results for each prompt alternative based on one or more performance criteria and automatically implementing one of the prompt alternatives based on the analysis of the recorded results.

Abstract: An interactive multimedia book provides hands-on multimedia instruction to the user in response to voiced commands. The book is implemented on a computer system and includes both text and audio/video clips. The interactive multimedia book is accessed by voiced commands and natural language queries as the primary user input. The displayed text is written in a markup language and contains hyperlinks which link the current topic with other related topics. The user may command the book to read the text and, as the text is read by the voice synthesizer, a word which is also a hyperlink will change its attributes upon being spoken. The user will be able to observe or hear this and simply utter the word which is the hyperlink to navigate to the linked topic.

Abstract: A system and method for switching between the command-mode of operation and the text-mode of operation in voice-recognition systems. The system operates using voice-recognition software and a mechanical switch which together enable the system to differentiate between spoken commands that are used to operate the system and spoken words that are inserted as text into a software application. The mechanical switch can be added to a conventional computer mouse, hand-held or desk-mounted microphone, personal digital assistant such as a Palm Pilot, or any other apparatus either used with or having a microphone. Alternatively, the system and method can be implemented using software alone without a mechanical switch, i.e., an electronic switch, wherein a specialized word or words are used to command the system to switch between the command-mode and the text-mode.

Abstract: The dynamic programming technique employs a lexical tree that is encoded in computer memory as a flat representation in which the nodes of each generation occupy contiguous memory locations. The traversal algorithm employs a set of traversal rules whereby nodes of a given generation are processed before the parent nodes of that generation. The deepest child generation is processed first and traversal among nodes of each generation proceeds in the same topological direction.

Abstract: A conversation system according to the invention includes: a network; a server apparatus which transfers a conversation based on characters through the network; plural terminal apparatuses which perform the conversation based on the characters under the transfer by the server apparatus through the network; and a translating apparatus which is connected to the network, receives a conversation that is transmitted from one of the plural terminal apparatuses through the network and based on characters of one language, translates the received conversation based on the characters of one language into a conversation based on characters of another language, and transmits the translated conversation based on the characters of another language to the server apparatus. The server apparatus transfers the conversation that is received from the translating apparatus and based on the characters of another language to at least one terminal apparatus during a conversation among the plural terminal apparatuses.

Abstract: A method for generating localizable message catalogs for Java-based applications is disclosed. Message catalogs that are automatically flagged for what needs to be manually translated are generated from a given Java source code file, which can then be used for translation. ListResourceBundle data structures that are compatible with Java's internationalization model are also generated from the message catalogs that were previously generated and manually translated into desired local language(s). This provides a more efficient means of maintaining a language-specific version of Java software after if has been released.

Abstract: The present invention provides a method for improving the efficiency of parsing text. Aspects of the invention include representing parse tokens as integers where a portion of the integer indicates the location in which a definition for the token can be found. In a further aspect, an integer representing a token points to an array of tokens that can be activated by the token. In another aspect, a list of pointers to partial parses is created before attempting to parse a next word in the text string. The list of pointers includes pointers to partial parses that are expecting particular semantic tokens. A fourth aspect of the invention utilizes a data structure to list the semantic tokens that have been fully parsed for each span in the input text segment. When a token is fully parsed, the list is accessed to determine if the new token should be discarded.

Abstract: In a distributed voice recognition system, a back-end pattern matching unit 27 can be informed of voice activity detection information as developed through use of a back-end voice activity detector 25. Although no specific voice activity detection information is developed or forwarded by the front-end of the system, precursor information as developed at the back-end can be used by the voice activity detector to nevertheless ascertain with relative accuracy the presence or absence of voice in a given set of corresponding voice recognition features as developed by the front-end of the system.

Abstract: The invention concerns a method for boosting data transmission in a telecommunications system and a mobile communications system. The invention is characterised in that the telecommunications network uses at least on a part of the transmission path between the fixed station, e.g. a base transceiver station, and the transcoder unit speech coding at a lower transmission rate than the transmission rate of the speech coding used on the transmission path between the fixed station and the terminal equipment. The speech parameters received from the terminal equipment are converted for the speech coding method used on the transmission connection between the fixed station and the transcoder unit and vice versa. In the network transcoder unit it is possible to reconvert speech parameters received from the direction of the terminal equipment e.g. into speech parameters of the speech coding used on the transmission path between the terminal equipment and the fixed station.

Abstract: A method and system of mixing audios to convert a plurality of input voices into a single output voice is described. The system of mixing audios has a decoding device, an audio mixing device and a frame package unit. The input voices including a plurality of audio frames are partially decoded to acquire audio parameters of the input voices by the decoding device. One audio frame of the input voices is selected by the audio mixing device to obtain a target frame according to the audio parameters later. The target frame is then packaged so as to be identical to the original format of the input voices by the frame package unit.

Abstract: The present invention provides an improved method and apparatus for translating a source language to a target language. The invention uses placeables (e.g., proper nouns, titles and names, dates, times, units and measurements, numbers, formatting information, such as tags or escape sequences, styles, graphics, hyperlinks) to assist a translator by not having to retype information that does not need to be translated and to provide conversions to the target locale if necessary like for speeds.

Abstract: A harmonic generation circuit compares audio data supplied from the outside every sample, detects a top-peak and an under-peak of the audio data on the basis of the comparison outputs, and detects patterns of the comparison output between continuous top-peak and under-peak. The harmonic generation circuit forms an addition-subtraction data corresponding to harmonic depending on the patterns, and supplies the addition-subtraction data to an adder at a timing depending on the patterns. The adder performs an addition-subtraction process for adding/subtracting the addition-subtraction data formed by the harmonic generation circuit to/from the audio data supplied from the outside. In this manner, with a compact, simple, and low-price circuit arrangement for performing only addition-subtraction, a harmonic component can be added to original audio data, and audio data having a wide frequency band can be formed.

Abstract: An apparatus and a concomitant method for speech recognition. In one embodiment, a distributed speech recognition system provides speech-driven control and remote service access. The distributed speech recognition system comprises a client device and a central server, where the client device is equipped with two speech recognition modules: a foreground speech recognizer and a background speech recognizer. The foreground speech recognizer is implementing a particular spoken language application (SLA) to handle a particular task, whereas the background speech recognizer is monitoring a change in the topic and/or a change in the intent of the user. Upon detection of a change in topic or intent of the user, the background speech recognizer will effect the routing to a new SLA to address the new topic or intent.

Abstract: A method and system for evaluating telephone services provided by speech recognition interfaces an evaluation engine with a voice recognition service over a telephone system to submit speech utterance samples to the voice recognition service, receive the response of the voice recognition service to the sample utterances, and determine error and recognition of the sample utterances by the voice recognition service by comparing actual voice recognition service responses to expected responses. The evaluation engine permits evaluation of a voice recognition service for plural glossaries in different contexts, such as through predetermined nodes of a voice recognition service menu having plural glossaries.

Abstract: A method and system, which may be implemented by employing a program storage device readable by machine, and tangibly embodying a program of instructions executable by the machine to perform method steps for ensuring robustness of a natural language understanding (NLU) system, includes tagging recognized words of a command input to the NLU system to associate the command with a context, and translating the command to at least one formal command based on the tagged words. A top ranked formal command is determined based on scoring of the tagged recognized words and scoring translations of the at least one formal command. Whether the top ranked formal command is accepted is determined by comparing a feature vector of the top ranked formal command to representations of feature vectors stored in an accept model. The top ranked formal command is executed if accepted and incorrect commands are prevented from execution to provide a robust NLU system.

Abstract: A system and method for providing a transcription service over a network. The transcription service includes a transcription server, a database, and several communication gateways. A transcription request is received by the system from anyone and includes an audio file, transcription instructions are derived by the system from a requestor profile, from preferences submitted with the audio file, or both. The request is then processed by the transcription server, the audio file is transmitted to at least one transcriber for transcription, transcribers receive the audio file, transcribe the file, and send the transcription back to the transcription server. The transcription server processes the file and sends the transcription to the entity that requested the transcription. Transcription accuracy is validated statistically by the system, thereby enabling reliable transcription from unknown transcription nodes. Profiles may be configured for both transcribers and transcription requestors.

Abstract: Methods and apparatus to process online interactions from numerous sources, coming from different time frames and systems. One embodiment includes a facilitator, an interaction processor, and an action generator. The facilitator facilitates interactions in one or more devices, where interactions typically occur. The processor can consolidate and analyze interactions from multiple sources, in different formats, and collected at different time frames, to extract intelligence from them. The processor processes such interactions, and generates an interaction descriptor for each interaction. A descriptor can include a generalization of the corresponding interaction, and at least a part of the interaction. Descriptors for different types of interactions can be represented by the same format to allow the processor to analyze them together.

Abstract: The present invention provides a speech feature extraction system suitable for use in a speech recognition system or other voice processing system that extracts features related to the frequency and amplitude characteristics of an input speech signal using a plurality of complex band pass filters and processing the outputs of adjacent band pass filters. The band pass filters can be arranged according to linear, logarithmic or mel-scales, or a combination thereof.