Abstract: In some implementations, a language proficiency of a user of a client device is determined by one or more computers. The one or more computers then determines a text segment for output by a text-to-speech module based on the determined language proficiency of the user. After determining the text segment for output, the one or more computers generates audio data including a synthesized utterance of the text segment. The audio data including the synthesized utterance of the text segment is then provided to the client device for output.

Abstract: Disclosed are a speech recognition verification device and a speech recognition verification method, which verify speech recognition results by executing artificial intelligence (AI) algorithms and/or machine learning algorithms in a 5G environment connected for Internet-of-Things. According to an embodiment, the speech recognition verification method includes converting a verification target text item to a verification target spoken utterance by applying a preset utterance condition, analyzing the verification target spoken utterance and outputting a recognition result text item corresponding to an analysis result, and verifying speech recognition performance through comparison between the verification target text item and the recognition result text item. According to the present disclosure, the speech recognition result may be verified objectively by using a spoken utterance generated with random text and various utterance conditions as input of speech recognition.

Abstract: A word extraction method according to at least one embodiment of the present disclosure includes: converting, with at least one processor operating with a memory device in a device, received speech information into text data; converting the text data into a string of words including a plurality of words; extracting, with the at least one processor operating with the memory device in the device, a keyword included in a keyword database from the plurality of words; and calculating, with the at least one processor operating with the memory device in the device, importance levels of the plurality of words based on timing of utterance of the keyword and timing of utterance of each of the plurality of words.

Abstract: A speech wakeup method, apparatus, and electronic device are disclosed in embodiments of this specification. The method includes: inputting speech data to a speech wakeup model trained with general speech data; and outputting, by the speech wakeup model, a result for determining whether to execute speech wakeup, wherein the speech wakeup model includes a Deep Neural Network (DNN) and a Connectionist Temporal Classifier (CTC).

Abstract: Disclosed are a speech data based language modeling system and method. The speech data based language modeling method includes transcription of text data, and generation of a regional dialect corpus based on the text data and regional dialect-containing speech data and generation of an acoustic model and a language model using the regional dialect corpus. The generation of an acoustic model and a language model is performed by machine learning of an artificial intelligence (AI) algorithm using speech data and marking of word spacing of a regional dialect sentence using a speech data tag. A user is able to use a regional dialect speech recognition service which is improved using 5G mobile communication technologies of eMBB, URLLC, or mMTC.

Abstract: According to an embodiment of the present disclosure, a method of updating a speech recognition model using a mobile agent in real-time comprises obtaining, in real-time, space type information for a particular space where the mobile agent is located, varying, in real-time, parameters of a speech recognition model used in the particular space based on the space type information, and performing a speech recognition service based on the speech recognition model including the varied parameters. Embodiments of the present disclosure may be related to artificial intelligence (AI) devices, unmanned aerial vehicles (UAVs), robots, augmented reality (AR) devices, virtual reality (VR) devices, and 5G service-related devices.

Abstract: A computer selects a test set of sentences from among sentences applied to train a whole sentence recurrent neural network language model to estimate the probability of likelihood of each whole sentence processed by natural language processing being correct. The computer generates imposter sentences from among the test set of sentences by substituting one word in each sentence of the test set of sentences. The computer generates, through the whole sentence recurrent neural network language model, a first score for each sentence of the test set of sentences and at least one additional score for each of the imposter sentences. The computer evaluates an accuracy of the natural language processing system in performing sequential classification tasks based on an accuracy value of the first score in reflecting a correct sentence and the at least one additional score in reflecting an incorrect sentence.

Abstract: Disclosed herein is a method for intelligently recognizing voice by a voice recognizing apparatus in various noise environments. The method includes acquiring a first noise level for an environment in which the voice recognizing apparatus is located, inputting the first noise level into a previously learned noise-sensitivity model to acquire a first optimum sensitivity, and recognizing a user's voice based on the first optimum sensitivity. The noise-sensitivity model is learned in a plurality of noise environments acquiring different noise levels, so that it is possible to accurately acquire an optimum sensitivity corresponding to a noise level depending on an operating state when an IoT device (voice recognizing apparatus) is in operation.

Abstract: Disclosed are an intelligent voice recognizing method, a voice recognizing device, and an intelligent computing device. According to an embodiment of the present invention, a method of intelligently recognizing a voice by a voice recognizing device obtains a microphone detection signal via at least one microphone, removes noise from the microphone detection signal based on a noise removal model, recognizes a voice from the noise-removed microphone detection signal, and updates the noise removal model based on the type of the noise detected from the microphone detection signal, thereby preventing deterioration of speech recognition performance. According to the present invention, one or more of the voice recognizing device, intelligent computing device, and server may be related to artificial intelligence (AI) modules, unmanned aerial vehicles (UAVs), robots, augmented reality (AR) devices, virtual reality (VR) devices, and 5G service-related devices.

Abstract: A dialogue system for a vehicle may include: an input processor configured to receive a dialogue among occupants of the vehicle including a driver and at least one passenger, to detect vehicle operation information, to identify the at least one passenger based on the dialogue among the occupants or the vehicle operation information, to generate passenger number information which estimates a change in a number of passengers in the vehicle based on the dialogue among the occupants when the vehicle arrives at a stop-over point, and to acquire a pre-utterance message according to the passenger number information; and a result processor configured to output a pre-utterance according to the pre-utterance message.

Abstract: A system that is capable of controlling multiple entertainment systems and/or speakers using voice commands. The system receives voice commands and may determine audio sources and speakers indicated by the voice commands. The system may generate audio data from the audio sources and may send the audio data to the speakers using multiple interfaces. For example, the system may send the audio data directly to the speakers using a network address, may send the audio data to the speakers via a voice-enabled device or may send the audio data to the speakers via a speaker controller. The system may generate output zones including multiple speakers and may associate input devices with speakers within the output zones. For example, the system may receive a voice command from an input device in an output zone and may reduce output audio generated by speakers in the output zone.

Abstract: Input context for a statistical dialog manager may be provided. Upon receiving a spoken query from a user, the query may be categorized according to at least one context clue. The spoken query may then be converted to text according to a statistical dialog manager associated with the category of the query and a response to the spoken query may be provided to the user.

Abstract: A method and an apparatus for generating information are provided. The method includes: determining, in response to receiving a first user sentence, whether a keyword of a preset first category is included in the first user sentence, the first category including at least one subcategory; determining, in response to determining the first category keyword being included in the first user sentence, the first category keyword included in the first user sentence as a first keyword, and determining a subcategory to which the first keyword belongs, to generate a first keyword set and a subcategory set; and selecting, based on the first keyword set and the subcategory set, a song list from a pre-generated song list set as a to-be-played song list, to generate a to-be-played song list set, the song list including at least one piece of audio and song list category information.

Abstract: Embodiments of the present disclosure disclose an interaction method and apparatus. A specific embodiment of the method includes: generating, in response to determining that a request input by a user satisfies a guiding condition, guiding information, and feeding back the guiding information to the user, the guiding condition including one of the following: associating with a plurality of query intents, or associating with no query intent; and generating, based on the request and a feedback input by the user corresponding to the guiding information, an intent-clear request, and feeding back push information bound with the intent-clear request to the user. Realizing that in the process of interacting with the user, for conditions such as the request input by the user is associated with a plurality of query intents or incompleteness, an intent-clear request associated with an explicit query intent is determined through the interaction with the user.

Abstract: An information processing device according to an aspect of the present technology includes a user information acquiring unit, an object information acquiring unit, and an output control unit. The user information acquiring unit acquires information related to a gaze position of a user while a substance of content is being automatically reproduced, in accordance with a first control amount, from an audio source located in a space in which the user is located. The object information acquiring unit acquires position information related to the audio source and position information related to a first object gazed at by the user. The output control unit performs first output control of providing the user with the substance of the content in accordance with a second control amount different from the first control amount in a case where the gaze position within the first object moves toward the audio source.

Abstract: The present technology relates to an information processing apparatus, an information processing method, and a program for enabling a message to be more reliably conveyed to a user. Provided with a presentation unit configured to present information to a first user, a detection unit configured to detect a reaction indicating that the first user has received the information, a search unit configured to search for a second user in a case where the detection unit has not been able to detect the reaction, and a request unit configured to request the second user found by the search unit to convey a message to the first user. A response promotion message asking for a response is output to the first user in the case where the detection unit has not been able to detect the reaction, and the search unit searches for the second user in the case where the detection unit has not been able to detect the reaction after the response promotion message has been output. The present technology can be applied to an agent device.

Abstract: It is an object of the present invention to promote a user's understanding or agreement, and to cause a dialogue to last long. A dialogue system 100 conducts a dialogue with a user 101. A humanoid robot 50-1 presents a first utterance which is a certain utterance. When the user 101 performs an action indicating that the user cannot understand the first utterance or it is predicted that the user performs an action indicating that the user cannot understand the first utterance or when the user does not perform any action indicating that the user can understand the first utterance, or it is predicted that the user will not perform any action indicating that the user can understand the first utterance, then the humanoid robot 50-1 presents a second utterance which is at least one utterance resulting from paraphrasing the contents of the first utterance.

Abstract: It is an object of the present invention to induce a dialogue to a topic that a dialogue system tries to present. A dialogue system 100 presents a first utterance which is a certain utterance and a target utterance related to the first utterance to a user 101. A humanoid robot 50-1 presents the first utterance. A microphone 11-1 receives a user utterance of the user 101 after the first utterance. A humanoid robot 50-2 presents at least one topic-inducing utterance for inducing the topic to the target utterance based on a recognition result of the user utterance and an utterance sentence of the target utterance after the user utterance. The humanoid robot 50-1 presents the target utterance after the topic-inducing utterance.

Abstract: A wireless event notification system includes a microphone, a controller, a voice authentication engine, a speech recognition engine, and a wireless device. The microphone is configured to transmit an audible command initiated by a user. The controller is located in a cloud and is configured to receive the audible command. The voice authentication engine is configured to receive the audible command for authenticating the user, and send an authentication signal to the controller. The speech recognition engine is configured to receive the audible command for recognition of the audible command, and send a command text indicative of the audible command to the controller. The wireless device is configured to receive the command text if the controller has received both the authentication signal from the voice authentication engine and the command text from the speech recognition engine.

Abstract: A control method for a human-computer interaction device, a human-computer interaction device, and a human-computer interaction system are described. The control method includes: capturing first voice information of a first object; identifying a second object related to the first voice information; acquiring first information related to the second object; and presenting the first information.

Abstract: Disclosed are a speech recognition method and a speech recognition device, in which speech recognition is performed by executing an artificial intelligence (AI) algorithm and/or a machine learning algorithm provided therein. According to one embodiment, the speech recognition method includes buffering a spoken utterance, extracting a standby wake-up word corresponding to a preset wake-up word from the spoken utterance by comparing the buffered spoken utterance to the preset wake-up word, analyzing the role of the standby wake-up word in the spoken utterance, determining the speech intent in uttering the standby wake-up word by using results of analyzing the role of the standby wake-up word, and determining whether to execute a spoken sentence as a voice command in the spoken utterance and processing the spoken sentence accordingly.

Abstract: Symbol sequences are estimated using a computer-implemented method including detecting one or more candidates of a target symbol sequence from a speech-to-text data, extracting a related portion of each candidate from the speech-to-text data, detecting repetition of at least a partial sequence of each candidate within the related portion of the corresponding candidate, labeling the detected repetition with a repetition indication, and estimating whether each candidate is the target symbol sequence, using the corresponding related portion including the repetition indication of each of the candidates.

Abstract: A speaker retrieval device includes a first converting unit, a receiving unit, and a searching unit. The first converting unit converts, using an inverse transform model of a first conversion model for converting score vectors representing the features of voice quality into acoustic models, pre-registered acoustic models into score vectors; and registers the score vectors in a corresponding manner to a speaker identifier in score management information. The receiving unit receives input of a score vector. The searching unit searches the score management information for the speaker identifiers whose score vectors are similar to the received score vector.

Abstract: A system and method for assisting communication through predictive speech is provided. A database includes commonly used words, phrases, and images, each associated with at least one context cue. A processor is configured to determine the user's context and display a number of possible initial words, phrases, or images associated with the determined context. A text field is updated with selected words, phrases, or images. The words, phrases, or literal equivalents of the images are audibly transmitted.

Abstract: A system for artificial intelligent dispute resolution is disclosed. The system may receive a dispute initiation request from a voice input channel. The system may determine user authentication state in response to the dispute initiation request. The system may receive a natural language problem statement from the voice input channel. The system may determine a user intent in response to the natural language problem statement. The system may compare the user intent with a business rules set and determine a dispositioned outcome based on the business rules set and the user intent.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for identifying a user in a multi-user environment. One of the methods includes receiving, by a first user device, an audio signal encoding an utterance, obtaining, by the first user device, a first speaker model for a first user of the first user device, obtaining, by the first user device for a second user of a second user device that is co-located with the first user device, a second speaker model for the second user or a second score that indicates a respective likelihood that the utterance was spoken by the second user, and determining, by the first user device, that the utterance was spoken by the first user using (i) the first speaker model and the second speaker model or (ii) the first speaker model and the second score.

Abstract: An error-concealing audio decoding method comprises: receiving a packet comprising a set of MDCT coefficients encoding a frame of time-domain samples of an audio signal; identifying the received packet as erroneous; generating estimated MDCT coefficients to replace the set of MDCT coefficients of the erroneous packet, based on corresponding MDCT coefficients associated with a received packet directly preceding the erroneous packet; assigning signs of a first subset of MDCT coefficients of the estimated MDCT coefficients, wherein the first subset comprises such MDCT coefficients that are associated with tonal-like spectral bins, to coincide with signs of corresponding MDCT coefficients of said preceding packet; randomly assigning signs of a second subset of MDCT coefficients of the estimated MDCT coefficients, wherein the second subset comprises MDCT coefficients associated with noise-like spectral bins; replacing the erroneous packet by a concealment packet containing the estimated MDCT coefficients and the s

Abstract: In general, techniques are described by which to embed enhanced audio transports in backward compatible bitstreams. A device comprising a memory and one or more processors may be configured to perform the techniques. The memory may store the backward compatible bitstream, which conforms to a legacy transport format. The processor(s) may obtain, from the backward compatible bitstream, legacy audio data that conforms to a legacy audio format, and obtain, from the backward compatible bitstream, extended audio data that enhances the legacy audio data. The processor(s) may also obtain, based on the legacy audio data and the extended audio data, enhanced audio data that conforms to an enhanced audio format, and output the enhanced audio data to one or more speakers.

Abstract: The present disclosure provides methods, devices and computer program products for encoding and decoding of a vector of parameters in an audio coding system. The disclosure further relates to a method and apparatus for reconstructing an audio object in an audio decoding system. According to the disclosure, a modulo differential approach for coding and encoding a vector of a non-periodic quantity may improve the coding efficiency and provide encoders and decoders with less memory requirements. Moreover, an efficient method for encoding and decoding a sparse matrix is provided.

Abstract: A method includes decoding a low-band portion of an encoded mid channel to generate a decoded low-band mid channel. The method also includes filtering the decoded low-band mid channel according to one or more filter coefficients to generate a low-band filtered mid channel. The method also includes generating an inter-channel predicted signal based on the low-band filtered mid channel and the inter-channel prediction gain. The method further includes generating a low-band left channel and a low-band right channel based on an up-mix factor, the decoded low-band mid channel, and the inter-channel predicted signal.

Abstract: The invention provides a decoder being configured for processing an encoded audio bitstream, wherein the decoder includes: a bitstream decoder configured to derive a decoded audio signal from the bitstream, wherein the decoded audio signal includes at least one decoded frame; a noise estimation device configured to produce a noise estimation signal containing an estimation of the level and/or the spectral shape of a noise in the decoded audio signal; a comfort noise generating device configured to derive a comfort noise signal from the noise estimation signal; and a combiner configured to combine the decoded frame of the decoded audio signal and the comfort noise signal in order to obtain an audio output signal.

Abstract: Methods, apparatus and articles of manufacture for research data gathering are disclosed. An example apparatus disclosed herein is to detect whether the apparatus is powered by an internal power source or an external power source. The example apparatus is also to, in response to detecting the apparatus is powered by the internal power source, perform first processing on a received audio signal to determine audio data to store in storage of the apparatus. The example apparatus is further to, in response to detecting the apparatus is powered by the external power source, perform second processing on the stored audio data to recover the code, the second processing different from the first processing.

Abstract: Example embodiments disclosed herein relate to signal processing. A method for decomposing a plurality of audio signals from at least two different channels is disclosed. The method comprises obtaining a set of components that are weakly correlated, the set of components generated based on the plurality of audio signals. The method comprises extracting a feature from the set of components, and determining a set of gains associated with the set of components at least in part based on the extracted feature, each of the gains indicating a proportion of a diffuse part in the associated component. The method further comprises decomposing the plurality of audio signals by applying the set of gains to the set of components. Corresponding system and computer program product are also disclosed.

Abstract: In one example, an apparatus includes: a wavelet transform engine to receive a first signal stream and perform a wavelet transform on a first time domain sample of the first signal stream, the first wavelet transform engine to output at least one first coefficient for a first frequency range; an energy calculation circuit to compute a first energy signature for the at least one first coefficient; and a correlation circuit to generate a correlation value using the first energy signature, a second energy signature and a plurality of previous energy signatures.

Abstract: What is described is an apparatus for post-processing an audio signal, having: a time-spectrum-converter for converting the audio signal into a spectral representation having a sequence of spectral frames; a prediction analyzer for calculating prediction filter data for a prediction over frequency within a spectral frame; a shaping filter controlled by the prediction filter data for shaping the spectral frame to enhance a transient portion within the spectral frame; and a spectrum-time-converter for converting a sequence of spectral frames having a shaped spectral frame into a time domain.

Abstract: A system for morphing an audio track includes a processor and software running on the processor. The software obtains target audio containing voice samples of a target voice and the software analyzes the target audio to create a target library. After the software creates the target library, the software loads a source audio file and, using the target library, the software morphs a voice from the source audio file into a morphed voice of the target voice, replacing the voice from the source file with the morphed voice of the target voice, creating a morphed audio file. The software then saves the morphed audio file into a storage associated with the processor.

Abstract: Methods and apparatuses for noise management are disclosed. In one example, a method includes receiving a plurality of noise level measurements. The method includes receiving a plurality of location data. In one example, the method further includes adjusting an environmental parameter utilizing the noise level measurements. In one example, the method further includes providing location services to a user directing the user to a geographical area having a lower noise level.

Abstract: In some embodiments, a method, apparatus and computer program for reducing noise from an audio signal captured by a drone (e.g., canceling the noise signature of a drone from the audio signal) using a model of noise emitted by the drone's propulsion system set, where the propulsion system set includes one or more propulsion systems, each of the propulsion systems including an electric motor, and wherein the noise reduction is performed in response to voltage data indicative of instantaneous voltage supplied to each electric motor of the propulsion system set. In some other embodiments, a method, apparatus and computer program for generating a noise model by determining the noise signature of at least one drone based upon a database of noise signals corresponding to at least one propulsion system and canceling the noise signature of the drone in an audio signal based upon the noise model.

Abstract: Noise filtering for an incoming signal is provided. The noise filtering method includes executing a transformation operation on the incoming signal by distributing energy corresponding to each of a plurality of components of the incoming signal into a two-dimensional representation. The noise filtering method also includes executing a filtering operation on the plurality of components to determine real objects and remove noise within the incoming signal. The filtering operation utilizing at least one of a plurality of noise detection matrixes based on time, frequency, or direction.

Abstract: In general, techniques are described by which to synchronize enhanced audio transports with backward compatible audio transports. A device comprising a memory and one or more processors may be configured to perform the techniques. The memory may store a backward compatible bitstream conforming to a legacy transport format. The processor may obtain, from the backward compatible bitstream, a first audio transport stream, and obtain, from the backward compatible bitstream, a second audio transport stream. The processor(s) may also obtain, from the backward compatible bitstream, indications representative of synchronization information for the first audio transport stream and the second audio transport stream. The processor(s) may synchronize, based on the indications, the first audio transport stream and the second audio transport to obtain synchronized audio data stream.

Abstract: A method for identifying at least one characteristic of a sound-producing object includes storing, in a memory, audio data acquired from an auditory environment via at least one microphone; receiving an input indicating a user request to identify a characteristic of a sound-producing object included in the auditory environment; determining, via a processor and based on a portion of the audio data acquired from the auditory environment prior to the user request, the characteristic of the sound-producing object; and causing information corresponding to the characteristic of the sound-producing object to be output via at least one output device.

Abstract: An emotion estimation system includes a feature amount extraction unit, a vowel section specification unit, and an estimation unit. The feature amount extraction unit analyzes recorded produced speech to extract a predetermined feature amount. The vowel section specification unit specifies, based on the feature amount extracted by the feature amount extraction unit, a section in which a vowel is produced. The estimation unit estimates, based on the feature amount in a vowel section specified by the vowel section specification unit, an emotion of a speaker.

Abstract: A spin transfer torque (STT) device is formed on an electrically conductive substrate and includes a ferromagnetic free layer near the substrate, a ferromagnetic polarizing layer and a nonmagnetic spacer layer between the free layer and the polarizing layer. A multilayer structure is located between the substrate and the free layer. The multilayer structure includes a metal or metal alloy seed layer for the free layer and an intermediate oxide layer below and in contact with the seed layer. The intermediate oxide layer reflects spin current from the free layer and thus reduces undesirable damping of the oscillation of the free layer's magnetization by the seed layer.

Abstract: A recording surface of a magnetic disk is divided into first and second zones. A first head of a first actuator arm assembly reads from and/or writes to the first zone exclusively. A second head of a second actuator arm assembly reads from and/or writes to the second zone exclusively. The first and second head are capable of simultaneously reading from and writing to the recording surface.

Abstract: A configuration is realized in which block encryption MMT format data is reproduced by applying a time stamp. An MMT format stream file and a reproduction control information file are generated and recorded in a medium. The stream file includes encryption block data to which an encryption key generated by using an additional header in which copy control information of a block unit is stored, as a seed, is applied, and the additional header. Position identification information capable of determining a position of reproduction data, a position of the seed to be applied to decoding of the reproduction data, a position of the time stamp, and a position of the seed to be applied to decoding of the time stamp is recorded in the reproduction control information file. Data decoding, and reproduction to which the time stamp is applied are performed by using recording information.

Abstract: The present invention enables correct placement of an electronic mark on a frame of captured image data intended by a remote control apparatus performing monitoring. Monitoring image data with a time code is transmitted to an external device. A command (an electronic mark placement command, or the like) to which a time code value showing a command target frame in the monitoring image data is added is received from the remote control apparatus. Processing based on the command is performed on a frame corresponding to the time code value added to the command, among image data corresponding to the monitoring image data recorded on a recording medium.

Abstract: The present disclosure provides a hard disk assembly device for assembling a hard disk into a case. Two opposite sides of the hard disk respectively have at least one first positioning portion and at least one second positioning portion. The hard disk assembly device includes a flexible fixing frame and a fixing bracket. One side of the flexible fixing frame includes at least one third positioning portion and at least one first guiding portion, and the other side of the flexible fixing frame includes at least one fourth positioning portion, a draw tape, and two fastening portions. The two fastening portions are respectively correspondingly connected with two ends of the draw tape. The hard disk is fixed in the flexible fixing frame. The flexible fixing frame is fixed to the fixing bracket.

Abstract: A non-volatile storage apparatus comprises a non-volatile memory structure and a plurality of I/O pads in communication with the non-volatile memory structure. The I/O pads include a power I/O pad, a ground I/O pad and data/control I/O pads. The non-volatile storage apparatus further comprises one or more capacitors connected to the power I/O pad. The one or more capacitors are positioned in one or more metal interconnect layers below the I/O pads.

Abstract: A semiconductor memory device includes n interconnect layers above a substrate; and a first interconnect region between an end of a control circuit and an end of the substrate in a direction of a first axis beside a first pad region in a direction of a second axis. The n interconnect layers are located at different levels from the substrate. Each of the n interconnect layers includes an interconnect. The first interconnect region includes no transistor, and no contact coupled to the substrate. The first interconnect region includes an interconnect extending along the second axis in m (m is a natural number equal to or larger than 3, larger than n/2, and equal to or smaller than n) interconnect layers of the n interconnect layers.

Abstract: In a semiconductor integrated circuit employing power gating, a control input signal is propagated to one or more first power switches through a first propagation path and to one or more second power switches through a second propagation path. A restoration determination circuit receives a first signal of the first propagation path and a second signal of the second propagation path and generates a control output signal. When the control signal performs restoration transition, the restoration determination circuit causes the control output signal to perform the restoration transition in accordance with a later timing of timings of restoration transitions of the first and second signals.