Abstract: A signal decoding apparatus that can suppress any large unusual sounds to provide decoded signals of improved audibility even when the number of hierarchical layers to be used in the decoding process varies due to a packet loss or the like in communication utilizing a scalable encoding/decoding technique. In the signal decoding apparatus, a gain adjusting part (2308) adjusts, based on a control of a decoding control part (2301), the gain of a basic layer decoded signal outputted from a basic layer decoding part (2302). A gain adjusting part (2309) adjusts, based on a control of the decoding control part (2301), the gain of a first expansion layer decoded signal outputted from a first expansion layer decoding part (2303). A gain adjusting part (2310) adjusts, based on a control of the decoding control part (2301), the gain of a second expansion layer decoded signal outputted from a second expansion layer decoding part (2304).

Abstract: A method, system and computer readable device for recognizing a partial utterance in an automatic speech recognition (ASR) system where said method comprising the steps of, receiving, by a ASR recognition unit, an input signal representing a speech utterance or word and transcribing the input signal into text, interpreting, by a ASR interpreter unit, whether the text is either a positive or a negative match to a list of automated options by matching the text with a grammar or semantic database representing the list of automated options, wherein if the ASR interpreter unit results in said positive match proceeding to a next input signal and if the ASR interpreter unit results in said negative match rejecting the text as representing said partial utterance, and processing, by a linguistic filtering unit, the rejected text to derive a correct match between the rejected text and the grammar or semantic database.

Abstract: A method for time aligning audio signal, wherein one signal has been derived from the other or both have been derived from another signal, comprises deriving reduced-information characterizations of the audio signals, auditory scene analysis. The time offset of one characterization with respect to the other characterization is calculated and the temporal relationship of the audio signals with respect to each other is modified in response to the time offset such that the audio signals are coicident with each other. These principles may also be applied to a method for time aligning a video signal and an audio signal that will be subjected to differential time offsets.

Abstract: A portal system identifies a language in which content can be displayed to the user. Possible languages can be determined from one or more of identity information for the user, a container for the user's directory entry, and the user's location. The determined languages can be ranked, and the highest-ranked available language can be selected for content display to the user.

Abstract: A method and system of providing language interpretation between a first person and a second person in visual contact with one another. The method comprises receiving from the remote language interpreter an audio signal emitted by the interpreter and video images of the interpreter; outputting the audio signal emitted by the interpreter; displaying the video images of the interpreter on a first display viewable by the first person while maintaining visual contact with the second person; and displaying the video images of the interpreter on a second display viewable by the second person while maintaining visual contact with the first person. In the case of sign language, the use of separate displays allows a hearing person and a non-hearing person to face one another while also viewing the interpreter.

Abstract: A speech recognition system processes speech outputs from a user to select a list element from a list of list elements. The user speaks the speech output as a whole word and an initial character sequence of the desired list element. The whole word and individual character sequence components are processed separately. A sublist is generated in response to the individual character sequence. A list element is selected in response to the whole word and sublist.

Abstract: A voice control system is directed to a car navigation device that includes: a display unit; a microphone; a speech recognition unit for recognizing, as a command, a speech uttered by a user; and a navigation ECU to which the voice recognition unit outputs the command. When the user inputs the speech, auxiliary switches are shown on the display unit. When the user selects one of the auxiliary switches, respective functions that correspond to the auxiliary switches are executed. The functions include displaying a voice command list or guidance for voice input on the displaying unit or outputting, to the navigation ECU, a result of determination whether a certain function, such as setting of a destination, should be allowed to be executed, the user first being required to affirm or negate the recognized destination setting by voice or auxiliary switch.

Abstract: This patent describes a method and apparatus to automatically and accurately winnow down arbitrarily large amounts of electronic information created by a particular population of actors to only those subsets of particular interest by having a causal relationship, even when retrieved documents containing this information do not individually satisfy the search criteria used. An actor in this context is defined as any entity, single or aggregate, capable of creating, distributing, modifying, or receiving digital information. Once identified, this subset of information may, for example, be processed, analyzed, redacted, or destroyed, depending on the context of the system's use.

Abstract: An apparatus for signal processing based on an algorithm for representing harmonics in a fractal lattice. The apparatus includes a plurality of tuned segments, each tuned segment including a transceiver having an intrinsic resonant frequency the amplitude of the resonant frequency capable of being modified by either receiving an external input signal, or by internally generating a response to an applied feedback signal. A plurality of signal processing elements are arranged in an array pattern, the signal processing elements including at least one function selected from the group including buffers for storing information, a feedback device for generating a feedback signal, a controller for controlling an output signal, a connection circuit for connecting the plurality of tuned segments to signal processing elements, and a feedback connection circuit for conveying signals from the plurality of signal processing elements in the array to the tuned segments.

Abstract: A method for pre-processing speech, in particular for recognizing speech, including receiving a speech signal, separating a spectrum of said speech signal into a number of predetermined frequency sub-bands, analyzing said speech signal within each of said frequency sub-bands, generating respective band-dependent acoustic feature data for each of said respective frequency sub-bands, deriving band-dependent likelihoods for occurrences of speech elements or within said speech signal based on said band-dependent acoustic feature data, analyzing said speech signal within said spectrum, generating full-band acoustic feature data, which are at least in part representative for said speech signal with respect to said spectrum, deriving a full-band likelihood for occurrences of speech elements or of sequences thereof within said speech signal based on said full-band acoustic feature data, deriving an overall likelihood for occurrences of speech elements within said speech signal based on said band-dependent likelihoods

Abstract: A method and apparatus for shaping quantization noise generated when compressing audio data at a low bit rate is disclosed. A predetermined quantization noise threshold allowed during quantization of sampled audio data and quantization noise energy information of a quantized MDCT coefficient are received in all frequency bands of an audio frequency. The quantization noise energy of the quantized MDCT coefficient is attenuated in a predetermined number of frequency bands in which a difference between the predetermined quantization noise threshold and the quantization noise energy of the quantized MDCT coefficient is large.

Abstract: Disclosed is a system and method for generating a spoken dialog service from website data. Spoken dialog components typically include an automatic speech recognition module, a language understanding module, a dialog management module, a language generation module and a test-to-speech module. These components are capable of being automatically trained from processed website data. A website analyzer converts a website into structured text data set and a structured task knowledge base. The website analyzer further extracts linguistic items from the website data. The dialog components are automatically trained from the structured text data set, structured task knowledge base and linguistic items.

Abstract: To provide an acoustic model which can absorb the fluctuation of a phonemic environment in an interval longer than a syllable, with the number of parameters of the acoustic model suppressed to be small, a phoneme-connected syllable HMM/syllable-connected HMM set is generated in such a way that a phoneme-connected syllable HMM set corresponding to individual syllables is generated by combining phoneme HMMs. A preliminary experiment is conducted using the phoneme-connected syllable HMM set and training speech data. Any misrecognized syllable and the preceding syllable of the misrecognized syllable are checked using results of a preliminary experiment syllable label data. The combination between a correct answer syllable for the misrecognized syllable and the preceding syllable of the misrecognized syllable is extracted as a syllable connection. A syllable-connected HMM corresponding to this syllable connection is added into the phoneme-connected syllable HMM set.

Abstract: The present invention relates to an information extraction apparatus capable of analyzing an acoustic signal with accuracy and high efficiency. An amplitude analysis section 32 determines whether or not an attack or release is contained on the basis of an amplitude value for each small region of an input time-series signal. When it is determined that there is an attack or release, an analysis region setting section 33 sets the portion from an attack position to a release position as an analysis region. A frequency analysis section 34 analyzes the input time-series signal by generalized harmonic analysis and outputs extracted waveform information. An extracted waveform synthesis section 35 synthesizes the extracted waveform information and outputs the information to a time-series compensation section 36. The time-series compensation section 36 compensates the signal of the synthesized result with a signal outside the analysis region and outputs an extracted waveform time-series signal to a subtraction unit 37.

Abstract: A privacy apparatus adds a privacy sound into the environment, thereby confusing listeners as to which of the sounds is the real source. The privacy sound may be based on the speaker's own voice or may be based on another voice. At least one characteristic of the speaker (such as a characteristic of the speaker's speech) may be identified. The characteristic may then be used to access a database of the speaker's own voice or another's voice, and to form one or more voice streams to form the privacy sound. The privacy sound may thus permit disruption of the ability to understand the source speech of the user by eliminating segregation cues that the auditory system uses to interpret speech.

Abstract: Hybrid linear predictive speech coding system with phase alignment predictive quantization zero phase alignment of speech prior to waveform coding aligns synthesized speech frames of a waveform coder with frames synthesized with a parametric coder. Inter-frame interpolation of LP coefficients suppresses artifacts in resultant synthesized speech frames.

Abstract: A system and method for determining quality of written product reviews to distinguish the user reviews for further use or processing. In one embodiment, the opinion analyzer system includes a processor, a parser module, an prose analyzer module, a characteristic features database, and a language value rules database. In another embodiment, the method comprises the steps of parsing language of a portion of a user review, extracting characteristic feature from the user review, determining a quality parameter based on the extracted characteristic feature, determining a raw score based on the quality parameter, and determining quality of the user review based on the raw score.

Abstract: In general, the invention is directed to translation techniques for acronyms and ambiguities. In one embodiment, translation of a source text into a target language includes identifying a potential acronym in the source text and presenting a translated acronym to the user as a candidate for translation of the acronym. In another embodiment, a user's response with respect to a first instance of an ambiguity is used to translate a second instance of the ambiguity in a set of related documents without any further user response.

Abstract: A word prediction method that improves the precision accuracy, and a speech recognition method and an apparatus therefor are provided. For the prediction of a sixth word “?”, a partial analysis tree having a modification relationship with the sixth word is predicted. “sara-ni sho-senkyoku no” has two partial analysis trees, “sara-ni” and “sho-senkyoku no”. It is predicted that “sara-ni” does not have a modification relationship with the sixth word, and that “sho-senkyoku no” does. Then, “donyu”, which is the sixth word from “sho-senkyoku no”, is predicted. In this example, since “sara-ni” is not useful information for the prediction of “donyu”, it is preferable that “donyu” be predicted only by “sho-senkyoku no”.

Abstract: A system and method for predicting lexical stress is disclosed comprising a plurality of stress prediction models. In an embodiment of the invention, the stress prediction models are cascaded, i.e. one after another within the prediction system. In an embodiment of the invention, the models are cascaded in order of decreasing specificity and accuracy. There is also provided a method of generating a lexical stress prediction system. In an embodiment, the method of generation includes generating a plurality of models for use in the system. In an embodiment, the models correspond to some or all of the models described above in relation to the first aspect of the invention.