Abstract: A technique is provided that can reduce degradation of the sound quality due to a tandem connection of paired coding and decoding, and can reduce the operation processing amount and the required memory amount of a multipoint control unit. At a terminal of a communication network having a larger communication capacity in multipoint connection between terminals in a plurality of communication networks (e.g., a fixed phone line and a mobile phone line) having different communication capacities, a multichannel coding including a monaural coding scheme of the communication network having the smaller communication capacity is performed on the coding side, whereas decoding of a multichannel-coded code of one point, decoding of a monaural-coded code of one point, or decoding of a monaural-coded code of a plurality of points is performed on the decoding side in accordance with the input code.

Abstract: An information processing device includes processing circuitry to acquire object spatiotemporal information including spatiotemporal information indicating coordinates of objects in time and space and a name of each of the objects and to generate morphological analysis-undergone object spatiotemporal information by executing a morphological analysis as a process of analyzing the name of each of the objects included in the object spatiotemporal information into one or more words; to acquire morphological analysis-undergone names of vicinal objects, as objects existing in a vicinity of each of the objects in time and space, from the morphological analysis-undergone object spatiotemporal information; to calculate a distribution of vicinal object name words, as words included in the names of the vicinal objects of each of the objects, from the morphological analysis-undergone names; and to convert the distribution of the vicinal object name words to a spatiotemporal information-considered distributed representati

Abstract: Systems and processes for providing user-specific acoustic models are provided. In accordance with one example, a method includes, at an electronic device having one or more processors, receiving a plurality of speech inputs, each of the speech inputs associated with a same user of the electronic device; providing each of the plurality of speech inputs to a user-independent acoustic model, the user-independent acoustic model providing a plurality of speech results based on the plurality of speech inputs; initiating a user-specific acoustic model on the electronic device; and adjusting the user-specific acoustic model based on the plurality of speech inputs and the plurality of speech results.

Abstract: In a method to decode signals, a computing device decodes spectral coefficients of a current frame are grouped into a plurality of sub-bands. The computing device classifies a sub-band as a bit allocation unsaturated sub-band based on an average quantity of allocated bits per spectral coefficient of a sub-band of the plurality of sub-bands and a threshold. The computing device obtains a noise filling gain based on an envelope of the sub-band, and obtains a reconstructed spectral coefficient of the sub-band by performing noise filling based on the noise filling gain. The computing device then obtains a frequency domain audio signal based on spectral coefficients in the sub-band obtained by decoding and the reconstructed spectral coefficient.

Abstract: A machine reading comprehension (MRC) question and answer providing method includes receiving a user question; analyzing the user question; selecting at least one document from at least one domain corresponding to an analyzed user question and searching for a passage, which is a candidate answer determined as being suitable for the user question, in the selected at least one document; obtaining at least one correct answer candidate value by inputting the user question and a corresponding passage into each of at least one MRC question and answer unit; and determining whether the at least one correct answer candidate value is a best answer.

Abstract: Machine classifiers in accordance with embodiments of the invention capture long-term temporal dependencies in particular tasks, such as turn-based dialogues. Machine classifiers may be used to help users to perform tasks indicated by the user. When a user utterance is received, natural language processing techniques may be used to understand the user's intent. Templates may be determined based on the user's intent in the generation of responses to solicit information from the user. A variety of persona attributes may be determined for a user. The persona attributes may be determined based on the user's utterances and/or provided as metadata included with the user's utterances. A response persona may be used to generate responses to the user's utterances such that the generated responses match a tone appropriate to the task. A response persona may be used to generate templates to solicit additional information and/or generate responses appropriate to the task.

Abstract: Systems and methods to group terms based on context to facilitate determining intent of a command are disclosed. Exemplary implementations to train a model: obtain a set of writings within a particular knowledge domain; obtain a vector generation model that generates vectors for individual instances of the terms in the set of writings; generate a first set of vectors that represent the instances of a first term and other vectors that represent instances of the other terms of the set of writings; train the vector generation model to group the vectors of a similar context in a space of a vector space; obtain a transcript include a new term generated from user audio dictation; generate a new vector that represent the instance of the new term; obtain the space; compare the new vector with the space; utilize the new term as the first term.

Abstract: A computerized method is disclosed including operations of receiving a plurality of request texts, and for each request text of the plurality of request texts: performing a pre-processing operation, performing a first text similarity procedure or a second text similarity procedure that each result in a determination of a most similar request text in a knowledge base, wherein the second text similarity procedure includes performance of word embedding operations, determining a degree of similarity between the current request text and the most similar request text, when the degree of similarity satisfies a similarity threshold comparison, associating an answer associated with the most request text with the current text request, and when the degree of similar does not satisfy the similarity threshold comparison, flagging the current request text or associating a placeholder answer with the current request text. Performing pre-processing may include removing stop words and punctuation and creating tokenized text.

Abstract: An encoding system encodes an N-channel audio signal (X), wherein N?3, as a single-channel downmix signal (Y) together with dry and wet upmix parameters ({tilde over (C)}, {tilde over (P)}). In a decoding system, a decorrelating section outputs, based on the downmix signal, an (N?1)-channel decorrelated signal (Z); a dry upmix section maps the downmix signal linearly in accordance with dry upmix coefficients (C) determined based on the dry upmix parameters; a wet upmix section populates an intermediate matrix based on the wet upmix parameters and knowing that the intermediate matrix belongs to a predefined matrix class, obtains wet upmix coefficients (P) by multiplying the intermediate matrix by a predefined matrix, and maps the decorrelated signal linearly in accordance with the wet upmix coefficients; and a combining section combines outputs from the upmix sections to obtain a reconstructed signal ({circumflex over (X)}) corresponding to the signal to be reconstructed.

Abstract: A device may receive a command associated with identifying a merchant for a virtual card swap procedure wherein the virtual card swap procedure is to replace a credit card of a user with a virtual card corresponding to the credit card. The device may identify the merchant for the virtual card swap procedure based on the command. The device may obtain the virtual card for the user. The device may determine a virtual card swap procedure template for the merchant. The device may perform the virtual card swap procedure based on the virtual card swap procedure template.

Abstract: A method of encoding an audio signal is provided comprising: applying multiple different time-frequency transformations to an audio signal frame; computing measures of coding efficiency across multiple frequency bands for multiple time-frequency resolutions; selecting a combination of time-frequency resolutions to represent the frame at each of the multiple frequency bands based at least in part upon the computed measures of coding efficiency; determining a window size and a corresponding transform size; determining a modification transformation; windowing the frame using the determined window size; transforming the windowed frame using the determined transform size; modifying a time-frequency resolution within a frequency band of the transform of the windowed frame using the determined modification transformation.

Abstract: Logic may adjust communications between customers. Logic may cluster customers into a first group associated with a first subset of synonyms and a second group associated with a second subset of the synonyms. Logic may associate a first tag with the first group and with each of the synonyms of the first subset. Logic may associate a second tag with the second group and with each of the synonyms of the second subset. Logic may associate one or more models with pairs of the groups. A first pair may comprise the first group and the second group. The first model associated with the first pair may adjust words in communications between the first group and the second group, based on the synonyms associated with the first pair, by replacement of words in a communication between customers of the first subset and customers of the second sub set.

Abstract: The present disclosure relates to techniques for identifying out-of-domain utterances.

Abstract: The disclosed subject matter relates to a system and method for providing an automated assistant that retrieves information from a knowledge base in response to a user's natural language communications. A refinement signal based on a subsequent user communication relevant to a previous search, establishes the type of search to be performed by the automated assistant in generating a response to the user communications. A reply communication to the user includes the selected results from search type selected. The types of searches are SPECIFIC, RELATIVE and NEGATION searches.

Abstract: A voice question and answer method and device, a computer readable storage medium, and an electronic device are described. The method comprises: receiving question voice information, and obtaining question text information according to the question voice information; performing at least one of general semantic analysis processing and dedicated semantic analysis processing on the question text information to generate an analysis result; and obtaining answer information according to the analysis result. The general semantic analysis is used for semantic analysis in the general field, and the dedicated semantic analysis is used for semantic analysis in the art field.

Abstract: Systems, devices, and methods are described for reducing degradation of a voice recognition input. An always listening device may always be listening for voice commands via a microphone and may experience interference from unwanted audio such as from the output audio of television speakers. The always listening device may receive data associated with the output audio over a first communications channel. The always listening device may also receive, on a second communications channel, timing information associated with data. The always listening device may adjust admission of the audio received by the microphone to enable it to arrive at approximately the same time as the data received via the first communications channel. The unwanted output audio included in the audio received via the microphone may then be determined and may be removed so that a voice command in the audio received by the microphone may be processed.

Abstract: A language device is described. A language device includes processing circuitry configured to: determine an urgency level of a message based at least in part on a natural language processing (NLP) model; determine a priority of an event associated with the message based at least in part on the urgency level; and order the message relative to a plurality of other messages in a customer queue based at least in part on the priority of the event associated with the message.

Abstract: A dialog apparatus and the like are provided which achieve a natural dialog in which topics minutely connect, starting from a user utterance, by making, after a robot's response to the user utterance, robots exchange additional questions and answers, which reflect the contents of the response, between them.

Abstract: Disclosed is a system and associated methods to optimize search in order to present emerging concepts that relate to a query for an established concept. The system extracts phrases from different content, generates a first cluster with a first set of phrases that contain a first common set of words and a second cluster with a second set of phrases that contain a second common set of words. The system clusters the first cluster and the second cluster to a cluster group based on a related word amongst the first and second set of phrases. The system receives a query for a particular word in the first common set of words, selects the cluster group based on the phrases in the first cluster of the cluster group having the particular word, and presents search results based on content that includes phrases in the second cluster of the cluster group.

Abstract: The present invention discloses a method and apparatus for awakening skills by speech, which are applied to an electronic device. The method for awakening skills by speech includes: recognizing awakening text information corresponding to a speech request message to be processed; invoking a service skill semantic model to determine a target service field corresponding to the awakening text information and a corresponding first confidence, and invoking a knowledge skill semantic model to determine a knowledge reply answer corresponding to the awakening text information and a corresponding second confidence; and selecting to awaken one of a knowledge skill and a target service skill corresponding to the target service field based on the first confidence and the second confidence. Accordingly, the probability of erroneously awakening a skill based on the speech message can be reduced.