Abstract: A translation training device which extracts from two nonparallel Corpora a set of parallel sentences. The system finds parameters between different sentences or phrases, in order to find parallel sentences. The parallel sentences are then used for training a data-driven machine translation system. The process can be applied repetitively until sufficient data is collected or until the performance of the translation system stops improving.

Abstract: A method of enabling the selection of a language to be employed as a method input language by a disambiguation routine of a handheld electronic device having stored therein a plurality of method input languages and disambiguation routine number, includes detecting a selection of a language, detecting as an ambiguous input an actuation of one or more input members, outputting at least a plurality of the language objects that corresponds to the ambiguous input, outputting an indicator which one of the method input languages is currently employed by the disambiguation function, and enabling an alternate one of the input language methods to be selected in response to a selection of the indicator in lieu of one of the plurality of language objects.

Abstract: The application discloses a communication device and method of processing a text message in the communication device. An aspect of the present application is a method of processing text message in a communication device, the method including receiving a text message from an external sender, determining whether and how a user desires to transform the text message into voice data according to user control information set by the user, selectively, transforming the received text message into voice data according to the result of the determining step, and transmitting the voice data to an external sound reproduction device through a wireless communication module.

Abstract: A personality-based theme may be provided. A prompt and an input indicating a personality may be received. Next, a voice font corresponding to the personality may be determined. The voice font may then be applied to the received prompt by augmenting the voice font applied prompt with recorded phrases of the personality.

Abstract: Generally methods for titling segments of recorded audio data are disclosed herein. An input from a voice activation module, a push button input or another user interface can provide a stimulus for a system or device to record title information. The title information can be received as an utterance, converted to text, and linked to a segment or body of recorded audio. A speech to text converter can perform the conversion from audio to text and the text can be displayed to a user. Then, the system can request and accept a confirmation from the user that the title information reflects a user's desires. In a recording retrieval mode, the system can display a plurality of titles with textual characters that represent a lingual translation of the title to the user and prompt the user for a user selection of a title. After such a selection is made, the recorded audio can be retrieved from memory and played back to the user over speakers or headphones.

Abstract: An input device for a mobile body allows for a safe input operation at the time of operating equipment, such as a car, regardless of whether the mobile body is traveling or stopped. The input device includes an input section for receiving inputs by hand and voice operations, a traveling state determination section for determining a traveling state of the mobile body, and an item executability determination section for determining, in the determined traveling state, whether each of the received inputs corresponds to an item executable by the hand and voice operations. The input device also includes a display content generation section for generating a display mode which distinguishably shows whether nor not the item is executable by the hand and voice operations in accordance with a determination by the item executability determination section, and a display section for displaying the generated display mode.

Abstract: Methods, and corresponding codec-containing devices are provided that have source coding schemes for encoding a component of an excitation. In some cases, the source coding scheme is an enumerative source coding scheme, while in other cases the source coding scheme is an arithmetic source coding scheme. In some cases, the source coding schemes are applied to encode a fixed codebook component of the excitation for a codec employing codebook excited linear prediction, for example an AMR-WB (Adaptive Multi-Rate-Wideband) speech codec.

Abstract: A system and method for managing localization workflow for video game development is disclosed. The method includes determining that a string is to be translated. The method further includes associating a translator with the string. The method further includes providing video metadata to the translator, wherein the video metadata includes at least video information from game play of the video game being developed such that the translator is provided with context for the translation.

Abstract: A process and system including a device, such as a set-top box, including a controller to detect a plurality of users engaging in a voice conference to discuss a presentation of a media program, convert speech dialog detected in the voice conference to textual dialog, detect from the textual dialog a behavioral profile of at least one of the plurality of users, and identify at least one of advertisement content and marketable media content for the plurality of users based on the behavioral profile of the at least one user. Other embodiments are disclosed.

Abstract: Techniques for assessing pronunciation abilities of a user are provided. The techniques include recording a sentence spoken by a user, performing a classification of the spoken sentence, wherein the classification is performed with respect to at least one N-ordered class, and wherein the spoken sentence is represented by a set of at least one acoustic feature extracted from the spoken sentence, and determining a score based on the classification, wherein the score is used to determine an optimal set of at least one question to assess pronunciation ability of the user without human intervention.

Abstract: The present invention discloses a payment card that uses speaker identification and verification (SIV) speech processing techniques for activation purposes and to deactivate the payment card based upon the card user input to the payment card. For example, the invention can initially identify a payment card in a deactivated state, which is an internal state of the payment card. Speech input can then be received. Speech characteristics of the speech input can be determined and compared against a voice print of an authorized card user. The payment card can be selectively activated based on comparison results. That is, when the voice print and the speech characteristics match, the payment card can be activated. Otherwise, the card will remain deactivated. An activated payment card is one that has undergone an internal state change from the deactivated state.

Abstract: A method for the artificial extension of the bandwidth of speech signals involves: a) Provision of a wideband input speech signal (swbi(k)); b) Determination of the signal components (seb(k)) of the wideband input speech signal (swbi(k)) required for the bandwidth extension from an extension band from the wideband input speech signal (swbi(k)); c) Determination of the temporal envelopes of the signal components (seb(k)) determined for the bandwidth extension; d) Determination of the spectral envelopes of the signal components (seb(k)) determined for bandwidth extension; e) Encoding of the information for the temporal envelopes and the spectral envelopes, and provision of the encoded information by carrying out the extension of the bandwidth; f) Decoding of the encoded information and generation of the temporal envelopes and the spectral envelopes from the encoded information for the production of a bandwidth-extended output speech signal (swbo(k)).

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for enhancing speech recognition accuracy. In one aspect, a method includes receiving geotagged audio signals that correspond to environmental audio recorded by multiple mobile devices in multiple geographic locations, receiving an audio signal that corresponds to an utterance recorded by a particular mobile device, determining a particular geographic location associated with the particular mobile device, generating a noise model for the particular geographic location using a subset of the geotagged audio signals, where noise compensation is performed on the audio signal that corresponds to the utterance using the noise model that has been generated for the particular geographic location.

Abstract: An enhancement system improves the estimate of noise from a received signal. The system includes a spectrum monitor that divides a portion of the signal at more than one frequency resolution. Adaptation logic derives a noise adaptation factor of the received signal. A plurality of devices tracks the characteristics of an estimated noise in the received signal and modifies multiple noise adaptation rates. Weighting logic applies the modified noise adaptation rates derived from the signal divided at a first frequency resolution to the signal divided at a second frequency resolution.

Abstract: Grammars for interactive voice response systems using natural language understanding can be created using information which is available on websites. These grammars can be created in automated manners and can have various tuning measures applied to obtain optimal results when deployed in a customer contact environment. These grammars can allow a variety of statements to be appropriately handled by the system.

Abstract: A wideband speech coding apparatus comprises a wideband speech coding unit configured to code an input speech signal at a bit rate which is set in advance in accordance with a wideband speech signal. The apparatus further comprises an identification unit configured to identify whether the input speech signal is a wideband speech signal or a narrowband speech signal. The apparatus further comprises a control unit configured to cause the wideband speech coding unit to code the input speech signal, when the identification unit identifies that the input speech signal is the wideband speech signal, and to cause the wideband speech coding unit to raise the bit rate to code the input speech signal, when the identification unit identifies that the input speech signal is the narrowband speech signal. The apparatus further comprises an output unit to output the input speech signal coded by the wideband speech coding unit.

Abstract: A noise elimination method and apparatus. The method eliminates noise from an input signal containing a voice signal mixed with a noise signal. The method includes detecting a noise section, in which the noise signal is present, from the input signal; obtaining a weight to be used for the input signal from signals of the noise section; and filtering the input signal using the obtained weight. The method and apparatus enable a mobile robot to eliminate noise in real time and effectively detect and recognize voice.

Abstract: An audio/music decoding device capable of improving quality of a decoded signal generated by conceal processing of a frame erase using a scalable encoding method. The audio/music decoding device includes a frame loss detector that determines whether encoded information is normally received and generates frame loss information indicating the result of the determination. According to the frame loss information, a first decoder performs decoding by using at least one of the following encoded information: the first encoded information on the frame immediately before, the first encoded information on the current frame, and the second encoded information on the current frame.

Abstract: Provided are, among other things, systems, methods and techniques for merging entropy codebook application ranges within an audio signal. According to one embodiment, an audio signal is obtained, the audio signal including quantization indexes, identification of segments of said quantization indexes, and indexes of entropy codebooks that have been assigned to such segments, with a single entropy codebook index having been assigned to each said segment; potential merging operations in which specified segments potentially would be merged with each other are identified; bit penalties are estimated for the potential merging operations; then, the potential merging operation having the lowest estimated the penalty is performed.

Abstract: In one embodiment, the method includes receiving an audio data frame having at least first and second channels. The first and second channels have been independently subdivided into blocks if the first and second channels are not correlated with each other, and the first and second channels have been synchronously subdivided into blocks if the first and second channels are correlated with each other and difference coding is used. The embodiment further includes obtaining subdivision information from the audio data frame. The subdivision information includes first information and second information. The first information indicates whether the first and second channels are independently subdivided or synchronously subdivided, and the second information indicates how the subdividing is performed. The first and second channels are decoded based on the obtained subdivision information.