Abstract: A method for adapting a speech recognition system for out-of-vocabulary, comprising, decoding by a hybrid speech recognition a speech including out-of-vocabulary terms, thereby generating graphemic transcriptions of the speech with a mixture of recognized in-vocabulary words and unrecognized sub-words, while keeping a track of the decoded segments of the speech, determining in the transcription sequences of sub-words as candidate out-of-vocabulary words based on a first condition with respect to lengths of the sequences of sub-words and a second condition with respect to the number of repetitions of the sequences, audibly presenting to a user the candidate out-of-vocabulary words from the corresponding segments of the speech according to the track, and receiving from the user indications of valid words corresponding to audible presentations of the sequences of sub-words in the candidate out-of-vocabulary words, and training a speech recognition to additionally recognize the candidate out-of-vocabulary words,

Abstract: A speech processing device includes an automotive device that filters data that is sent and received across an in-vehicle bus. The device selectively acquires vehicle data related to a user settings or adjustments. An interface acquires the selected vehicle data from in-vehicle sensors in response to a user's articulation of a first code phrase. A memory stores the selected vehicle data with unique identifying data associated with the user and establishes a connection between the selected vehicle data and the user when a second code phrase is articulated. A data interface provides access to the selected vehicle data and stored relationship data and enables the processing of the data to customize the in-vehicle system. The data interface is responsive to the user's articulation of a third code phrase to process the selected vehicle data that enables the setting or adjustment of the in-vehicle system.

Abstract: A navigation apparatus capable of providing a user not only with guidance, but also with all of the guidance, operational procedure, operation screen and recognition vocabulary, that is, with an operational transition that is defined by the guidance, operational procedure, operation screen and recognition vocabulary, while altering the operational transition in accordance with the recognition vocabulary comprehension level of the user. Thus, it can increase the possibility for a user with a low recognition vocabulary comprehension level to achieve a task, or for a user with a high recognition vocabulary comprehension level to improve the comfortableness of the operation, thereby being able to provide all the users with the optimum operational transition.

Abstract: Noise reduction system with remote noise detector The present invention relates to a noise reduction system with at least one remote noise detector placed close to at least one noise source, which transmits relevant information to a primary device where it is used for noise reduction. Thereby, acoustic signal enhancement can be achieved via the at least one remote noise detector in that a noise estimate is transmitted to controller for noise reduction in the signal obtained from a primary source.

Abstract: Sound decomposition models are described. In one or more implementations, a plurality of individual models is generated for respective ones of a plurality of sound sources. The plurality of models is collected to form a universal audio model that is configured to support sound decomposition of sound data through use of one or more of the models. The plurality of models is not generated using a sound source that originated at least a portion of the sound data.

Abstract: Disclosed herein are systems, methods, and non-transitory computer-readable storage media for assigning saliency weights to words of an ASR model. The saliency values assigned to words within an ASR model are based on human perception judgments of previous transcripts. These saliency values are applied as weights to modify an ASR model such that the results of the weighted ASR model in converting a spoken document to a transcript provide a more accurate and useful transcription to the user.

Abstract: Methods and systems for receiving applications are described. A device, such as a wearable computing device, may receive an input, which may be verbal, motion, or text, for example. The device may parse the input to recognize any keywords and may further identify applications associated with the recognized keywords. The device may determine a priority list for the identified applications based at least in part on a relevance associated with the recognized keywords. In addition, the device may provide for display the priority list of the applications. The device may receive a request to access an application from the plurality of applications, and in response to the request to access the application, install the application. The device may further execute the installed application to perform a function based at least in part on the verbal input, which may be executed without additional input after installation.

Abstract: Embodiments described herein provide approaches for validating synonyms in ontology driven natural language processing. Specifically, an approach is provided for receiving a user input containing a token, structuring the user input into a semantic model comprising a set of classes each containing a set of related permutations of the token, designating the token as a synonym of one of the set of related permutations, annotating the token with a class from the set of classes corresponding to the one of the set of related permutations, and validating the annotation of the token by determining an accuracy of the designation of the token as a synonym of the one of the set of related permutations. In one embodiment, the accuracy is determined by quantifying a linear distance between the token and a contextual token also within the user input, and comparing the linear distance to a pre-specified linear distance limit.

Abstract: Provided are a coding device, a communication processing device, and a coding method, whereby processing operation load (computational load) is significantly reduced for a configuration which computes either frame energy or sub-frame energy of an input signal, using auto-correlation operations, without causing a decline in the precision of either the frame energy or the sub-frame energy. In a coding device (101), a sub-frame energy computation unit (201) computes the sub-frame energy by substituting the sum of input signal auto-correlation operations in a first range with the sum of auto-correlation operations in a second range which differs at least partially from the first range.

Abstract: Obtaining and/or validating user credentials at client devices is described. A phrase may be generated based on one or more index values determined according to a function of time and a credential identifier identifying a user credential. The phrase may be output by the client device for validating the user credential.

Abstract: A voice recognition device includes a vehicle state detecting unit for detecting a vehicle state of a vehicle having the voice recognition device mounted thereon, and an acoustic data matching unit for matching an acoustic feature value converted by a sound analyzing unit with a recognition dictionary stored in a recognition dictionary storage unit to recognize a voice input to a microphone. The device also includes a recognition parameter setting unit for setting a recognition parameter at the time when the voice input to the microphone is recognized, and a control unit for instructing the recognition parameter setting unit to change the recognition parameter when the vehicle state detected by the vehicle state detecting unit satisfies a predetermined condition.