Abstract: A method and system for concealing errors in one or more bad frames in a speech sequence as part of an encoded bit stream received in a decoder. When the speech sequence is voiced, the LTP-parameters in the bad frames are replaced by the corresponding parameters in the last frame. When the speech sequence is unvoiced, the LTP-parameters in the bad frames are replaced by values calculated based on the LTP history along with an adaptively-limited random term.

Abstract: A procedure for the automatic generation of a textual expression from a semantic representation by a computer-system is described. With the procedure, a statistical model is determined by the computer-system on a plurality of pre-determined pairs of semantic representations and associated expressions and stored. A semantic representation, from which an associated expression is determined by the computer system by means of the statistical model, is presented to the computer system. These steps are repeated by the computer system for further semantic representations if necessary.

Abstract: A voice controlled electronic device includes a controller (12, 13, 14) for initiating individual functions of the electronic device. The controller also establishes a language attribute associated with a language for interaction with the user. The controller ensures that at least part of the interaction with the user takes place substantially in the associated language. The electronic device includes an input (1) for receiving voice commands. A speech recognizer (4) recognizes at least one voice command in the speech input. The voice command is associated with a predetermined first control function of a device, and a distinct second function of establishing the language attribute. The controller sets the language attribute according to the second function of the recognized command.

Abstract: According to one embodiment of the invention, a multistage vector list quantizer comprises a first stage quantizer to select candidate first stage codewords from a plurality of first stage codewords, a reference table memory storing a set of second stage codewords for each first stage codeword, and a second stage codebook constructor to generate a reduced complexity second stage codebook that is the union of sets corresponding to the candidate first stage codewords selected by the first stage quantizer.

Abstract: Disclosed are scalable quantizers for audio and other signals characterized by a non-uniform, perception-based distortion metric, that operate in a common companded domain which includes both the base-layer and one or more enhancement-layers. The common companded domain is designed to permit use of the same unweighted MSE metric for optimal quantization parameter selection in multiple layers, exploiting the statistical dependence of the enhancement-layer signal on the quantization parameters used in the preceding layer. One embodiment features an asymptotically optimal entropy coded uniform scalar quantizer. Another embodiment is an improved bit rate scalable multi-layer Advanced Audio Coder (AAC) which extends the scalability of the asymptotically optimal entropy coded uniform scalar quantizer to systems with non-uniform base-layer quantization, selecting the enhancement-layer quantization methodology to be used in a particular band based on the preceding layer quantization coefficients.

Abstract: The invention concerns a system and method of predicting problematic dialogs in a task classification system based on the user's input communications. The method may include determining whether a task classification decision can be made based on a first automated dialog exchange with the user. As such, if the task classification decision cannot be made, the method may determine whether the probability of conducting a successful automated dialog with the user based on whether the first dialog exchange exceeds a first threshold. The successful dialog may be defined as a dialog exchange between an automated dialog system and the user that results in at least one of processing of the user's input communication and routing the user's input communication. The method may further operate such that if the first threshold is exceeded, further dialog is conducted with the user. Otherwise, the user may be directed to a human for assistance.

Abstract: In a mobile wireless communication system automatic speech recognition is performed in a distributed manner using a mobile station based near or front end stage which extracts and vector quantizes recognition feature parameters from frames of an utterance and an infrastructure based back or far end stage which reverses the vector quantization to recover the feature parameters and subjects the feature parameters to a Hidden Markov Model (HMM) evaluation to obtain a recognition decision for the utterance. In order to conserve network capacity, the size (Sz) of the codebook used for the vector quantization, and the corresponding number of bits (B) per codebook index B, are adapted on a dialogue-by dialogue basis in relation to the vocabulary size |V| for the dialogue. The adaptation is performed at the front end, accomplishes a tradeoff between expected recognition rate RR and expected bitrate BR by optimizing a metric which is a function of both.

Abstract: An information processing apparatus for separating input musical number information into a vocal information part containing lyrics in a first language and an accompaniment information part, and for producing second musical number information made of the accompaniment part and a translated vocal information part superimposed thereon. A vocal separation unit separates the first vocal information part and the accompaniment information part from the input first musical information. A processing unit generates first language lyric information by speech recognition of the separated first vocal information part, translates the generated first language lyric information into second language lyric information, and supplies the second language lyric information. A synthesis unit synthesizes the supplied second language lyric information, the accompaniment information part, and the separated first vocal information part to generate second musical information.

Abstract: Given a source text, a desired translation of the source text into a target language, and a machine-readable dictionary, a first set of morphemes in the target language is generated from the source text, typically by using the dictionary to perform a machine translation of the source text. The second text is analyzed into a second set of morphemes in the target language. Differences between the first and second sets of morphemes are found, and morphemes corresponding to the differences are taken from the source text. Existing information including these source-text morphemes is extracted from the dictionary, and new information to be added to the dictionary is automatically generated from the extracted information and the differences. This process generates comparatively short dictionary entries, corresponding only to the differences between the two set of morphemes, and therefore creates useful dictionary entries while saving dictionary space.

Abstract: In a digital channel of a digital wireless communication system including at least one mobile station, at least one base transceiver station in communication with the mobile station, a transcoder configured to provide a signal conversion between vocoder frames and pulse code modulation, and a mobile switching center for interconnecting the digital wireless communication system to a public switched telephone network, a method and apparatus for determining a fault in the digital channel is disclosed. The method includes generating a first set of vocoder input parameters from a speech input signal, and generating a second set of vocoder input parameters from an output signal substantially equivalent to the speech input signal as it is received at a mobile station via the digital channel. The method further includes calculating a metric based on the first and the second set of vocoder input parameters, and subsequently determining a fault in the digital channel using the metric.

Abstract: A translation device which has both advantages of a table look-up translation device and advantages of a machine translation device by leading the user's utterance through a sentence template suitable for the user's intent of speech is realized. Since the translation device searches for sentence templates suitable for the user's intent of speech with an orally inputted keyword and displays retrieved sentences, the user's utterance can be lead. In addition, the user is free from a troublesome manipulation for replacing a word since an expression uttered by the user is inserted into a replaceable portion (slot) within the sentence template, and the translation device translates a resulting sentence with the replaced expression embedded in the slot.

Abstract: With respect to audio signal coding and decoding apparatuses, there is provided a coding apparatus that enables a decoding apparatus to reproduce an audio signal even through it does not use all of data from the coding apparatus, and a decoding apparatus corresponding to the coding apparatus. A quantization unit constituting a coding apparatus includes a first sub-quantization unit comprising sub-quantization units for low-band, intermediate-band, and high-band; a second sub-quantization unit for quantizing quantization errors from the first sub-quantization unit; and a third sub-quantization unit for quantizing quantization errors which have been processed by the first sub-quantization unit and the second sub-quantization unit.

Abstract: There are provided speech coding methods and systems for estimating a plurality of speech parameters of a speech signal for coding the speech signal using one of a plurality of speech coding algorithms, the plurality of speech parameters includes pitch information, the plurality of speech parameters is calculated using a plurality of thresholds. An example method includes estimating a background noise level in the speech signal to determine a signal to noise ratio (SNR) for the speech signal, adjusting one or more of the plurality of thresholds based on the SNR to generate one or more SNR adjusted thresholds, analyzing the speech signal to extract the pitch information using the one or more SNR adjusted thresholds, and repeating the estimating, the adjusting and the analyzing to code the speech signal using one the plurality of speech coding algorithms.

Abstract: When a user's request is entered it is then transmitted to a network interface unit which digitizes and stores the request. The digitized request and information about the user's network of devices is then transmitted from the network interface unit to a natural language server, preferably over the internet. The natural language server then processes the request and generates commands necessary to complete the request within the user's network of devices. These commands are then transmitted from the natural language server to the network interface unit. The network interface unit then transmits the commands to the appropriate devices within the network of devices. The devices within the network of devices then execute the received commands to complete the user's request.

Abstract: A hash function based data retrieval system for use with a lexicon database of a data processing system is disclosed. The system comprises a RAM, a disk memory, and a data file residing in the disk memory, wherein the data file contains stored data and is organized into nests. The system further comprises a hashing data structure residing in RAM, wherein the data structure is designed to occupy a fixed amount of memory independent of content of the data file. A data retrieval module is operable to identify a nest using a hash function that is based on parameters selected according to characteristics of the data file. The hash function is further designed to be optimized for content of the data file. The hash function is further designed to produce hash values based on the fixed amount of memory.

Abstract: A common narrow-band speech signal is expanded into a wide-band speech signal. The expanded speech signal gives the impression of a wide-band speech signal regardless of what type of vocoder is used. Extending the narrow-band speech signal into a lower range involves analyzing the narrow-band speech signal to generate one or more parameters, and synthesizing a lower frequency-band signal based on at least one of the one or more parameters. The synthesized lower frequency-band signal is then combined with a signal that is derived from (e.g., via up-sampling) the narrow-band speech signal. In preferred embodiments, a pitch frequency parameter is generated, and generation of the lower frequency-band signal includes generating continuous sine tones that are frequency shifted with the pitch frequency parameter.

Abstract: A method and apparatus for detecting and transmitting voice signals in a packet voice network system. The method and apparatus make use of a voice activity detection (VAD) unit at a transmitter, for determining if an input signal contains active audio information or passive audio information, where the input signal includes a plurality of frames. For one or more frames of the input signal containing active audio information, the VAD computes a hangover time period. This computation includes determining whether the hangover time period has a fixed duration or a variable duration on the basis of characteristics of the active audio information contained in the one or more frames. When the VAD detects a frame containing passive audio information subsequent to the one or more frames containing active audio information, the input signal is suppressed after the expiry of the computed hangover time period from the detection of the passive audio information.

Abstract: Methods and apparatus for controlling an electronic device connected to a network are provided. The methods and apparatus described herein convert a text based device list and/or a text based function list into text based voice prompt scripts. The voice prompt scripts are then read to a user via a text-to-speech engine. The user responds with a voice command for a device. The voice command is converted to text by a voice recognition engine. This text is then used to send a command to the electronic device via the network.

Abstract: The present invention can be used in a natural language processing system to determine a relationship (such as similarity in meaning) between two textual segments. The relationship can be identified or determined based on logical graphs generated from the textual segments. A relationship between first and second logical graphs is determined. This is accomplished regardless of whether there is an exact match between the first and second logical graphs. In one embodiment, the first graph represents an input textual discourse unit. The second graph, in one embodiment, represents information in a lexical knowledge base (LKB). The input graph can be matched against the second graph, if they have similar meaning, even if the two differ lexically or structurally.

Abstract: The present invention relates to an algorithm for ensuring compliancy of an algorithm module when integrated in a real time software system. The compliancy tests may include a memory test, interrupt test, latency test and other tests, as well as combinations thereof. An inventive aspect of the present invention relates to a unit test harness for verifying that a software algorithm module meets performance and functional requirements when integrated in a complete real-time software system. A software algorithm module eliminates or reduces unwanted behavior by the caller or other software on a real-time software system due to incorrect operations, which may involve interrupts, memory usage, register usage and/or other factors.