Abstract: A technology of accurately coding and decoding coefficients which are convertible into linear prediction coefficients even for a frame in which the spectrum variation is great while suppressing an increase in the code amount as a whole is provided. A coding device includes: a first coding unit that obtains a first code by coding coefficients which are convertible into linear prediction coefficients of more than one order; and a second coding unit that obtains a second code by coding at least quantization errors of the first coding unit if (A-1) an index Q commensurate with how high the peak-to-valley height of a spectral envelope is, the spectral envelope corresponding to the coefficients which are convertible into the linear prediction coefficients of more than one order, is larger than or equal to a predetermined threshold value Th1 and/or (B-1) an index Q? commensurate with how short the peak-to-valley height of the spectral envelope is, is smaller than or equal to a predetermined threshold value Th1?.

Abstract: Methods and systems for recognizing and performing voice commands during advertisements are provided. An example method may include playing, by an audio device, a media stream to a user, the media stream including at least one advertisement; sensing, by an acoustic sensor, an ambient acoustic signal; determining, by processors communicatively coupled to the audio device and the acoustic sensor, that the audio device has started playing the advertisement; in response to the determination, monitoring, by the processors, the ambient acoustic signal to detect a presence of at least one command spoken by the user; and in response to the detection of the presence of the at least one command, determining data associated with the at least one advertisement; and causing, by the processors, the audio device to perform one or more actions associated with the command and the data associated with the advertisement.

Abstract: A third-party company may assist companies in providing natural language interfaces for their customers. To implement a natural language interface for a company, a configuration may be received that includes information, such as a list intents, seed messages for the intents, and hierarchical information of the intents. An intent classifier may be trained using the configuration, and the natural language interface may be deployed for use with customers. Usage data of the natural language classifier may be collected and used to improve the natural language interface. Messages corresponding to an intent may be clustered into clusters of similar messages, and a prototype message may be obtained for each cluster to provide a human understandable description of the cluster. The information about the clusters may be used to improve the natural language interface, such as by creating a new intent with a cluster or moving a cluster to a different intent.

Abstract: Systems, methods, and computer-readable storage devices are disclosed for generating smart notes for a meeting based on participant actions and machine learning. One method including: receiving meeting data from a plurality of participant devices participating in an online meeting; continuously generating text data based on the received audio data from each participant device of the plurality of participant devices; iteratively performing the following steps until receiving meeting data for the meeting has ended, the steps including: receiving an indication that a predefined action has occurred on the first participating device; generating a participant segment of the meeting data for at least the first participant device from a first predetermined time before when the predefined action occurred to when the predefined action occurred; determining whether the receiving meeting data of the meeting has ended; and generating a summary of the meeting.

Abstract: Natural speech dialog system and methods are disclosed. In one example, a method includes identifying a dialog system intent associated with the speech input based on at least one predetermined intent keyword, the dialog system intent having required intent parameters, determining whether data for all required intent parameters of the dialog system are available, based on the determination, selectively initiating a parameter collection dialog associated with the dialog system intent, the parameter collection dialog being operable to collect data for the required parameters not otherwise available to the dialog system intent, and based on the dialog system intent and one or more required parameters, generating an action instruction.

Abstract: A greeting card having an audio message recording and playback device permits recording of personalized audio messages to be played upon opening of the greeting card. The recording device is operable in either a trial mode or a use mode. In the trial mode, which would be applicable when the card is displayed in a store, a potential purchaser may experience the functionality of the card by recording their own test message. The test message is played back initially for the potential purchaser but is not subsequently played back to be later heard by other potential purchasers. In the use mode, which the card may be switched to after purchase by removal of a trial mode panel from the greeting card, a user recorded message is played repeatedly upon subsequent openings of the card. The user recorded message may be followed by a prerecorded recording, such as a song. Additional prerecorded messages, such as voice prompts with instructions for recording a message, may also be included.

Abstract: User generated content, particularly Chinese language content, is retrieved from various sources such as forums, microblogs, social media sites, and the like. A portion of the content is manually labeled with a sentiment associated with the content and may be classified according to subject matter referenced. Sentiment-indicating features of the content is extracted according to a sentiment dictionary, which may include topic-specific jargon. The features are used to train a classifier to determine sentiment of content based on sentiment-indicating features. The sentiment for other content may then be determined using the classifier. The output of the classifier may be combined with an explicit rating of a product or product feature.

Abstract: A method and a device for recognition, and a method and a device for constructing a recognition model are disclosed. A device for constructing a recognition model includes a training data inputter configured to receive additional training data, a model learner configured to train a first recognition model constructed based on basic training data to learn the additional training data, and a model constructor configured to construct a final recognition model by integrating the first recognition model with a second recognition model generated by the training of the first recognition model.

Abstract: There is provided an apparatus and a method for rapidly extracting a target sound from a sound signal where a variety of sounds are mixed generated from a plurality of the sound sources. There is a voice recognition unit including a tracking unit for detecting a sound source direction and a voice segment to execute a sound source extraction process, and a voice recognition unit for inputting a sound source extraction result to execute a voice recognition process. In the tracking unit, a segment being created management unit that creates and manages a voice segment per unit of sound source sequentially detects a sound source direction, sequentially updates a voice segment estimated by connecting a detection result to a time direction, creates an extraction filter for a sound source extraction after a predetermined time is elapsed, and sequentially creates a sound source extraction result by sequentially applying the extraction filter to an input voice signal.

Abstract: A networked communication system is described. The communication system including an automatic speech recognizer configured to receive a speech signal from a client over a network and to convert the speech signal into a text sequence. The communication also including a speech analyzer configured to receive the speech signal. The speech analyzer configured to extract paralinguistic characteristics from the speech signal. In addition, the communication system includes a speech output device coupled with the automatic speech recognizer and the speech analyzer. The speech output device configured to convert the text sequence into an output speech signal based on the extracted paralinguistic characteristics.

Abstract: A computer-implemented method is provided. The computer-implemented method is performed by a speech recognition system having at least a processor. The method includes estimating sound identification information from a neural network having periodic indications and components of a frequency spectrum of an audio signal data inputted thereto. The method further includes performing a speech recognition operation on the audio signal data to decode the audio signal data into a textual representation based on the estimated sound identification information. The neural network includes a plurality of fully-connected network layers having a first layer that includes a plurality of first nodes and a plurality of second nodes. The method further comprises training the neural network by initially isolating the periodic indications from the components of the frequency spectrum in the first layer by setting weights between the first nodes and a plurality of input nodes corresponding to the periodic indications to 0.

Abstract: There is provided a speech synthesizer comprising a processor configured to receive one or more linguistic units, convert said one or more linguistic units into a sequence of speech vectors for synthesizing speech, and output the sequence of speech vectors. Said conversion comprises modelling higher and lower spectral frequencies of the speech data as separate high and low spectral streams by applying a first set of one or more statistical models to the higher spectral frequencies and a second set of one or more statistical models to the lower spectral frequencies.

Abstract: Natural speech dialog system and methods are disclosed. In one example, a method includes identifying a dialog system intent associated with the speech input based on at least one predetermined intent keyword, the dialog system intent having required intent parameters, determining whether data for all required intent parameters of the dialog system are available, based on the determination, selectively initiating a parameter collection dialog associated with the dialog system intent, the parameter collection dialog being operable to collect data for the required parameters not otherwise available to the dialog system intent, and based on the dialog system intent and one or more required parameters, generating an action instruction.

Abstract: A voice recording device that connects/is connected to a network, comprising a voice recording circuit that acquires voice and records the acquired voice as a voice file, a transmission circuit that transmits the voice file to a network, and a control circuit, the control circuit including an information extraction section that extracts associated information that has been associated with the voice file, and a display that displays the associated information associated with a voice data file.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for presenting notifications in an enterprise system. In one aspect, a method include actions of obtaining enrollment acoustic data representing an enrollment utterance spoken by a user, obtaining a set of candidate acoustic data representing utterances spoken by other users, determining, for each candidate acoustic data of the set of candidate acoustic data, a similarity score that represents a similarity between the enrollment acoustic data and the candidate acoustic data, selecting a subset of candidate acoustic data from the set of candidate acoustic data based at least on the similarity scores, generating a detection model based on the subset of candidate acoustic data, and providing the detection model for use in detecting an utterance spoken by the user.

Abstract: Audio information defining audio content may be accessed. The audio content may have a duration. The audio content may be segmented into audio segments. Individual audio segments may correspond to a portion of the duration. The audio segments may include a first audio segment corresponding to a first portion of the duration. Energy features, entropy features, frequency features, and/or other features of the audio segments may be determined. Energy features may characterize energy of the audio segments. Entropy features may characterize spectral flatness of the audio segments. Frequency features may characterize highest frequencies of the audio segments. One or more of the audio segments may be identified as containing speech based on the energy features, the entropy features, the frequency features, and/or other information. Storage of the identification of the one or more of the audio segments as containing speech in one or more storage media may be effectuated.

Abstract: An audio coding terminal and method is provided. The terminal includes a coding mode setting unit to set an operation mode, from plural operation modes, for input audio coding by a codec, configured to code the input audio based on the set operation mode such that when the set operation mode is a high frame erasure rate (FER) mode the codec codes a current frame of the input audio according to a select frame erasure concealment (FEC) mode of one or mom FEC modes. Upon the setting of the operation mode to be the High FER mode, the one FEC mode is selected, from the one or more FED modes predetermined for the High FER mode, to control the codec by incorporating of redundancy within a coding of the input audio or as separate redundancy information separate from the coded input audio according to the selected one FEC mode.

Abstract: The present invention proposes a new method for improving the performance of a real-valued filterbank based spectral envelope adjuster. By adaptively locking the gain values for adjacent channels dependent on the sign of the channels, as defined in the application, reduced aliasing is achieved. Furthermore, the grouping of the channels during gain-calculation, gives an improved energy estimate of the real valued subband signals in the filterbank.

Abstract: A device for noise estimation comprises a first microphone capturing a nominal speech signal, and a second microphone capturing a nominal noise signal. A generalized sidelobe canceller of the device applies spatial noise reduction, and comprises a blocking matrix filter to adaptively process the nominal speech signal to produce a speech cancellation signal, a node for subtracting the speech cancellation signal from the nominal noise signal to produce a noise reference signal, a noise cancellation filter to adaptively filter the noise reference signal to produce a noise cancellation signal; and a node for subtracting the noise cancellation signal from the nominal speech signal to produce a speech reference signal.

Abstract: Various embodiments described herein facilitate multi-lingual communications. The systems and methods of some embodiments may enable multi-lingual communications through different modes of communications including, for example, Internet-based chat, e-mail, text-based mobile phone communications, postings to online forums, postings to online social media services, and the like. Certain embodiments may implement communications systems and methods that translate text between two or more languages (e.g., spoken), while handling/accommodating for one or more of the following in the text: specialized/domain-related jargon, abbreviations, acronyms, proper nouns, common nouns, diminutives, colloquial words or phrases, and profane words or phrases.