Abstract: A method on a mobile device for voice recognition training is described. A voice training mode is entered. A voice training sample for a user of the mobile device is recorded. The voice training mode is interrupted to enter a noise indicator mode based on a sample background noise level for the voice training sample and a sample background noise type for the voice training sample. The voice training mode is returned to from the noise indicator mode when the user provides a continuation input that indicates a current background noise level meets an indicator threshold value.

Abstract: The subject matter of this specification can be embodied in, among other things, a method that includes receiving geographical information derived from a non-verbal user action associated with a first computing device. The non-verbal user action implies an interest of a user in a geographic location. The method also includes identifying a grammar associated with the geographic location using the derived geographical information and outputting a grammar indicator for use in selecting the identified grammar for voice recognition processing of vocal input from the user.

Abstract: Techniques are disclosed relating to selecting and/or ranking from among multiple recordings based on determining facial attributes associated with detected words. For example, each recording may be transcribed and analyzed to determine whether the recordings include words in one or more sets of words. Facial analysis may be performed during intervals in the recordings corresponding to recognized words. Counts of the recognized words may be weighted based on detected facial attributes. In various embodiments, disclosed techniques may facilitate accurate selection of relevant media from large data sets.

Abstract: Techniques for using a dynamic wakeword detection threshold are described. A server(s) may receive audio data corresponding to an utterance from a device in response to the device detecting a wakeword using a wakeword detection threshold. The server(s) may then determine the device should use a lower wakeword detection threshold for a duration of time. In addition to sending the device output data responsive to the utterance, the server(s) may send the device an instruction to use the lower wakeword detection threshold for the duration of time. Alternatively, the server(s) may train a machine learning model to determine when the device should use a lower wakeword detection threshold. The server(s) may send the trained machine learned model to the device for use at runtime.

Abstract: A cooperative conversational voice user interface is provided. The cooperative conversational voice user interface may build upon short-term and long-term shared knowledge to generate one or more explicit and/or implicit hypotheses about an intent of a user utterance. The hypotheses may be ranked based on varying degrees of certainty, and an adaptive response may be generated for the user. Responses may be worded based on the degrees of certainty and to frame an appropriate domain for a subsequent utterance. In one implementation, misrecognitions may be tolerated, and conversational course may be corrected based on subsequent utterances and/or responses.

Abstract: Provided herein is a voice recognition server and a control method thereof, the method including determining an index value for each of a plurality of training texts; setting a group for each of the plurality of training texts based on the index values of the plurality of training texts, and matching a function corresponding to each group and storing the matched results; in response to receiving a user's uttered voice from a user terminal apparatus, determining an index value from the received uttered voice; and searching a group corresponding to the index value determined from the received uttered voice, and performing the function corresponding to the uttered voice, thereby providing a voice recognition result of a variety of user's uttered voices suitable to the user's intentions.

Abstract: Speech recognition methods, devices, and systems are described herein. One system includes a number of microphones configured to capture sound in an area, a digital signal processor configured to segregate the captured sound into a plurality of signals, wherein each respective signal corresponds to a different portion of the area, and an automatic speech recognition engine configured to separately process each of the plurality of signals to recognize a speech command in the captured sound.

Abstract: Method and systems are provided for interpreting speech data. A method and system for recognizing speech involving a filter module to generate a set of processed audio data based on raw audio data; a translation module to provide a set of translation results for the raw audio data; and a decision module to select the text data that represents the raw audio data. A method for minimizing noise in audio signals received by a microphone array is also described. A method and system of automatic entry of data into one or more data fields involving receiving a processed audio data; and operating a processing module to: search in a trigger dictionary for a field identifier that corresponds to the trigger identifier; identify a data field associated with a data field identifier corresponding to the field identifier; and providing content data associated with the trigger identifier to the identified data field.

Abstract: A hearing aid device for processing signals comprising audio in order to assist a hearing impaired user and being configured to be arranged at, behind and/or in an ear of the user is disclosed. The hearing aid device comprises at least one input transducer, an output transducer, an antenna, a wireless interface, and a signal processing unit. The at least one input transducer is adapted for receiving signals comprising audio. The output transducer is adapted for providing signals comprising audio to the user. The antenna is adapted for wirelessly communicating with a respective external device. The wireless interface is adapted for receiving and/or sending data via the antenna. The signal processing unit is adapted for processing signals comprising audio in order to assist a hearing impaired user. Furthermore the signal processing unit is adapted for extracting speech signals from the signals comprising audio and for generating control signals based on the speech signals.

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: A language storage method according to one aspect of the present invention includes vectorization steps for vectorizing language accumulated in a dialog log database using a word vector created in advance for each word; determination step for determining whether to employ the language that has been vectorized as response data; and a storage step for accumulating the language that has been determined to be employed as the response data in a response database, in which a dimension reduction step for reducing the number of dimensions of the vector of the language vectorized in the vectorization step is included between the vectorization step and the determination step, and in the determination step, it is determined which one of an employed group and an unemployed group the vector of the language whose number of dimensions has been reduced in the dimension reduction step belongs to.

Abstract: A method, computer system, and a computer program product for optimizing speech recognition in a smart medical room. The present invention may include receiving a piece of verbal data associated with a medical encounter from one or more audio recording devices. The present invention may also include accessing a plurality of signals from a plurality of biometric sensors associated with a plurality of medical equipment associated with the smart medical room based on the received piece of verbal data associated with the medical encounter. The present invention may further include selecting, from a database, one or more speech domain models based on the accessed plurality of signals from the plurality of biometric sensors associated with the plurality of medical equipment, wherein the one or more speech domain models are utilized to optimize a transcription of speech during the medical encounter in the smart medical room.

Abstract: Methods employ sensors in portable devices (e.g., smartphones) both to sense content information (e.g., audio and imagery) and context information. Device processing is desirably dependent on both. For example, some embodiments activate certain processor intensive operations (e.g., content recognition) based on classification of sensed content and context. The context can control the location where information produced from such operations is stored, or control an alert signal indicating, e.g., that sensed speech is being transcribed. Some arrangements post sensor data collected by one device to a cloud repository, for access and processing by other devices. Multiple devices can collaborate in collecting and processing data, to exploit advantages each may have (e.g., in location, processing ability, social network resources, etc.). A great many other features and arrangements are also detailed.

Abstract: One embodiment provides a method, including receiving, at an audio capture device, a customized activation cue; identifying, using a processor, contextual information associated with a user; analyzing, using the contextual information, characteristics of the customized activation cue; identifying, based on the analyzation, a uniqueness associated with the customized activation cue; and responsive to said identifying, notifying a user that the customized activation cue has inadequate uniqueness. Other aspects are described and claimed.

Abstract: A method of encoding samples in a digital signal is provided that includes receiving a frame of N samples of the digital signal, determining L possible distinct data values in the N samples, determining a reference data value in the L possible distinct data values and a coding order of L?1 remaining possible distinct data values, wherein each of the L?1 remaining possible distinct data values is mapped to a position in the coding order, decomposing the N samples into L?1 coding vectors based on the coding order, wherein each coding vector identifies the locations of one of the L?1 remaining possible distinct data values in the N samples, and encoding the L?1 coding vectors.

Abstract: Disclosed are various embodiments for detecting replay attacks in voice-based authentication systems. In one embodiment, audio is captured via an audio input device. It is then verified that the audio includes a voice authentication factor spoken by a user. The audio is then compared with stored audio spoken by the user. If it is determined that an exact copy of the voice authentication factor is in the stored audio, one or more actions may be performed.

Abstract: The present technology relates to an encoding device and a method, a decoding device and a method, and a program that enables acquisitions of high-quality sound even in a resource-poor setting. A demultiplexer demultiplexes a supplied code string, to obtain the quantized low-band spectrum, the spectral characteristic code, and the quantized expansion coefficient(s). At this point, the code string includes a single quantized expansion coefficient or quantized expansion coefficients of the respective bands in the high band depending on the spectral characteristic code. A spectral inverse quantization unit obtains the low-band spectrum by inversely quantizing the quantized low-band spectrum. An expansion coefficient inverse quantization unit obtains the expansion coefficient(s) by inversely quantizing the quantized expansion coefficient(s).

Abstract: A speech/audio coding apparatus includes a receiver that receives a time-domain speech input signal. The apparatus also includes a processor that transforms a time-domain speech input signal into a frequency-domain spectrum, and divides a frequency region of the spectrum in an extended band into a plurality of bands. The processor sets a limited band for each divided band in the current frame, a width of the limited band in the current frame being narrower than the divided band and the limited band including a first frequency. The processor further encodes the spectrum in the limited band within each divided band in the current frame, wherein the width of the limited band is predetermined and is set to 31.

Abstract: The present document relates to time-alignment of encoded data of an audio encoder with associated metadata, such as spectral band replication (SBR) metadata. An audio decoder (100, 300) configured to determine a reconstructed frame of an audio signal (237) from an access unit (110) of a received data stream is described. The access unit (110) comprises waveform data (111) and metadata (112), wherein the waveform data (111) and the metadata (112) are associated with the same reconstructed frame of the audio signal (127). The audio decoder (100, 300) comprises a waveform processing path (101, 102, 103, 104, 105) configured to generate a plurality of waveform subband signals (123) from the waveform data (111), and a metadata processing path (108, 109) configured to generate decoded metadata (128) from the metadata (111).

Abstract: A voice processing method and device, the method comprising: detecting a current voice application scenario in a network (S1); determining the voice quality requirement and the network requirement of the current voice application scenario (S2); based on the voice quality requirement and the network requirement, configuring voice processing parameters corresponding to the voice application scenario (S3); and according to the voice processing parameters, conducting voice processing on the voice signals collected in the voice application scenario (S4).

Abstract: A method is provided for resampling an audio-frequency signal in an audio-frequency signal encoding or decoding operation. The resampling is carried out by a method of interpolation of an order greater than one. The method is such that the interpolated samples are obtained by calculating a weighted average of possible interpolation values calculated over a plurality of intervals covering the time location of the sample to be interpolated. A resampling device is provided, which implements the method, and also an encoder and decoder including at least one resampling device.

Abstract: An approach to speech synthesis uses two phases in which a relatively low quality waveform is computed, and that waveform is passed through an enhancement phase which generates the waveform that is ultimately used to produce the acoustic signal provided to the user. For example, the first phase and the second phase are each implemented using a separate artificial neural network. The two phases may be computationally preferable to using a direct approach to yield a synthesized waveform of comparable quality.

Abstract: A command processing device and method are provided. The command processing device includes a receiving device and a processing device. The receiving device receives an audio signal from an electronic device, wherein the audio data includes data signals and clock signals, and the data signals correspond to a first sound channel and the clock signals correspond to a second sound channel. The processing device is coupled to the receiving device. The processing device obtains the data signals from the first sound channel, and obtains the clock signals from the second sound channel. The processing device obtains one or more commands according to the data signals and clock signals, and performs operations corresponding to the commands.

Abstract: Systems and methods for enhancing reverberated audio signals are disclosed. In one embodiment, a method is disclosed comprising receiving an audio signal; partitioning a frequency domain representation of the audio signal into a plurality of sub-band vectors; inputting each sub-band vector into a corresponding deep neural network; calculating, using the corresponding deep neural networks, a plurality of output vectors for each sub-band; concatenating the plurality of output vectors to generate a clean audio feature matrix; and converting the clean audio feature matrix into a time-domain audio signal.

Abstract: An audio processing apparatus is provided that includes an acquirer that acquires a surrounding audio signal indicating a sound surrounding a user. The audio processing apparatus also includes an audio extractor that extracts, from the acquired surrounding audio signal, a providing audio signal indicating a sound to be provided to the user. The audio processing apparatus further includes an output that outputs a first audio signal, indicating a main sound, and the providing audio signal.

Abstract: The technology described in this document can be embodied in a computer-implemented method that includes receiving information representing audio captured by a microphone array, wherein the information includes multiple datasets each representing audio signals captured in accordance with a sensitivity pattern along a corresponding direction with respect to the microphone array. The method also includes computing, using one or more processing devices for each of the multiple datasets, one or more quantities indicative of human voice activity captured from the corresponding direction, and generating, based at least on the one or more quantities computed for a plurality of the multiple datasets, a directional audio signal representing audio captured from a particular direction.

Abstract: A pitch detection method. Such a pitch detection method may apply Pseudo Weigner Ville Transformation (PWVT) as a spectral representation of speech signal. Also, the pitch detection method may take the median value of each frame of the speech signal as a threshold for making the voicing decision. Additionally, the pitch detection method may take a moving average of PWVT as the threshold for voicing decision.

Abstract: Devices that include a near field transducer (NFT) including a crystalline plasmonic material having crystal grains and grain boundaries; and nanoparticles disposed in the crystal grains, on the grain boundaries, or some combination thereof, wherein the nanoparticles are oxides of, lanthanum (La), barium (Ba), strontium (Sr), erbium (Er), hafnium (Hf), germanium (Ge), or combinations thereof; nitrides of zirconium (Zr), niobium (Nb), or combinations thereof; or carbides of silicon (Si), aluminum (Al), boron (B), zirconium (Zr), tungsten (W), titanium (Ti), niobium (Nb), or combinations thereof.

Abstract: A recording head includes a near-field transducer proximate a media-facing surface. The near-field transducer comprises an aperture portion surrounded by walls of plasmonic material, the walls oriented normal to the media-facing surface. A notch protrudes within the aperture. The notch comprises at least one of Rh and Ir. A write pole is proximate the near-field transducer. The write pole has a back surface facing away from the media-facing surface and an aperture-facing surface proximate the aperture.

Abstract: The magnetic tape device includes: a magnetic tape; and a servo head, in which a magnetic tape transportation speed of the magnetic tape device is equal to or lower than 18 m/sec, the servo head is a magnetic head including a tunnel magnetoresistance effect type element as a servo pattern reading element, the magnetic tape includes a non-magnetic support, and a magnetic layer including ferromagnetic powder and a binding agent on the non-magnetic support, the magnetic layer includes a servo pattern, and a coefficient of friction measured regarding a base portion of a surface of the magnetic layer is equal to or smaller than 0.30.

Abstract: A method of operating a storage device includes reading data from a storage medium using a detector, processing signals from the detector through a plurality of processing circuits, each respective processing circuit in the plurality of processing circuits being optimized for a different state of a channel condition and providing a respective output metric, selecting a processing circuit from the plurality of processing circuits by comparing the respective output metrics from each processing circuit in a predetermined manner, and designating as output of the detector output of the processing circuit that is selected. The output metrics may be branch metrics or path metrics, and the channel condition may be fly-height or phase shift. The storage device includes a storage medium, and a read channel including a detector, and processing circuits that process signals from the detector. Each respective processing circuit is optimized for a different state of a channel condition.

Abstract: A magnetic tape is provided in which the total thickness is less than or equal to 5.30 ?m. The magnetic layer includes a timing-based servo pattern and also a fatty acid ester; and the magnetic layer surface Ra is less than or equal to 1.8 nm. The full width at half maximum of spacing distribution measured by optical interferometry regarding the surface of the magnetic layer before performing vacuum heating with respect to the magnetic tape is greater than 0 nm and less than or equal to 7.0 nm, while the same value measured after performing the vacuum heating is greater than 0 nm and less than or equal to 7.0 nm, also. The difference between a spacing measured after performing the vacuum heating and a spacing measured before performing the vacuum heating is greater than 0 nm and less than or equal to 8.0 nm.

Abstract: A magnetic tape includes a non-magnetic support; a non-magnetic layer including non-magnetic powder and a binder on the non-magnetic support; and a magnetic layer including ferromagnetic powder and a binder on the non-magnetic layer. The total thickness of the non-magnetic layer and the magnetic layer is less than or equal to 0.60 ?m. The magnetic layer includes a timing-based servo pattern, the ferromagnetic powder is ferromagnetic hexagonal ferrite powder, and the magnetic layer includes an abrasive. In addition, the tilt cos ? of the ferromagnetic hexagonal ferrite powder with respect to the surface of the magnetic layer acquired by cross section observation performed using a scanning transmission electron microscope is 0.85 to 1.00.

Abstract: A magnetic tape is provided in which the total thickness of a non-magnetic layer and a magnetic layer is equal to or smaller than 0.60 ?m. The magnetic layer includes a timing-based servo pattern and also includes fatty acid ester. The full width at half maximum of spacing distribution measured by optical interferometry regarding the surface of the magnetic layer before performing vacuum heating with respect to the magnetic tape is greater than 0 nm and less than or equal to 7.0 nm, while the same value measured after performing the vacuum heating is also greater than 0 nm and less than or equal to 7.0 nm. Further, the difference between a spacing measured after performing the vacuum heating and a spacing measured before performing the vacuum heating is greater than 0 nm and equal to or smaller than 8.0 nm.

Abstract: A light source-unit comprises a laser diode, a sub-mount which the laser diode is joined, and a heater which is joined on a joint surface of the sub-mount. The sub-mount comprises a pair of barrier-members. The barrier-members are formed with lower thermal conductivity material which thermal conductivity is lower than joining metal which is used for joining the laser diode and the sub-mount. The barrier-members are formed on the joint surface so as to sandwich the heater.

Abstract: A tape library rack includes partitions defining a slot for receiving a tape cartridge. A hook is coupled with the rack to engage a notch in the cartridge and retain the cartridge within the rack slot. A gripper for retrieving the cartridge includes a ramped surface to engage the hook and laterally displace the hook from the cartridge notch as the hook travels along the ramped surface. The gripper engages the notch from which the hook has been displaced and pulls the cartridge from the slot.

Abstract: First tracks are read via a first head that is moved via a first actuator over a first, radially-defined, zone of a disk surface. Second tracks are read via a second head that is moved via a second actuator over a second zone of the disk surface that is separate from the first zone. The first and second heads are optimized to read data within first and second skew angle ranges associated with the first and second zones. The first and second skew angle ranges are each less than a total skew angle range of the disk surface.

Abstract: A storage system such as a hard disc drive (HDD), solid-state drive (SSD), hybrid drive (SSHD), storage rack, set of storage racks, JBOD, array of discs, etc. may include a variety of storage media. Failures may be detected in the storage media of the storage system. Such a failure may affect the physical capacity of the storage system. A storage controller of the storage system initiates a media conversion that converts a portion of the storage media from media storing data according to a first recording scheme to a media storing data according to a second recording scheme. The second recording scheme stores data at a higher density compared to the first recording scheme.

Abstract: The invention pertains in general to a latching mechanism for maintaining desired friction levels on an optical disc in an optical disc restoration device. In particular the invention pertains to devices, systems and methods for easily maintaining friction levels between pads and an optical disc in an optical disc restoration device for ease of adjusting friction settings during quality control, repair operation or when optimization settings are being set in an optical disc restoration device by a user.

Abstract: A movie editor converts a received movie into a proxy format, and creates a texture strip representing the frames of the movie. An editor can use the texture strip to edit the movie, rather than editing the movie directly. Deep tags and/or special effects can be defined for the texture strip using a graphical interface. The graphical interface enables movies to be combined into a playback product according to a control structure graphically presented in the graphical interface.

Abstract: A system and method for combining light codes and a video are provided. The system includes an editing device and a storage device. The editing device identifies the content of a video to obtain a playback period of a plurality of frames having at least one candidate object in the video. The storage device stores at least one candidate object code and at least one light code symbol, and the light code symbol has a correspondence with the candidate object code. The editing device matches playback period of the frames having the at least one candidate object with the light code symbol, so as to provide a light code output table corresponding to the video.

Abstract: Provided herein is an apparatus including a hard drive base. A number of disks are rotatably connected to the hard drive base, and there are a number of disk gaps between the disks. A shroud extends in a circumferential span around the disks, wherein the shroud extends less than the circumference of the disks. The shroud is positioned between the disk gaps and an upwardly sloping gas channel. A number of arcuate gas channels are in the shroud, wherein each arcuate gas channel is horizontally aligned with a corresponding disk gap. The arcuate gas channels connect the disk gaps to the upwardly sloping gas channel.

Abstract: A method for controlling operations of a memory device, the memory device and the controller thereof, and the associated electronic device are provided. The method can comprise: before a voltage-drop event regarding a driving voltage occurs, mapping a rising reference voltage and a falling reference voltage to a first reference voltage and a second reference voltage, respectively; when the voltage-drop event occurs, pausing at least one access operation to a non-volatile (NV) memory, and mapping the rising reference voltage and the falling reference voltage to another first reference voltage and another second reference voltage, respectively; and when the voltage-drop event ends, mapping the rising reference voltage and the falling reference voltage to the first reference voltage and the second reference voltage, respectively.

Abstract: An electronic device includes an internal supply rail; a plurality of first main header switches for coupling the internal supply rail to a first power supply; a plurality of second main header switches for coupling the internal supply rail to a second power supply; an auxiliary circuit including a first auxiliary header switch for coupling the internal supply rail to the first power supply and a second auxiliary header switch for coupling the internal supply rail to the second power supply; a feedback circuit, the feedback circuit tracking a status of the first and second main header switches; and a control circuit, the control circuit controlling the first main header switches, second main header switches and first and second auxiliary header switches responsive to the switch control signal and an output of the feedback circuit.

Abstract: The present disclosure includes apparatuses and methods for data transfer between subarrays in memory. An example may include a first subarray of memory cells and a second subarray of memory cells, wherein a first portion of memory cells of the first subarray and a first portion of memory cells of the second subarray are coupled to a first sensing circuitry stripe. A third subarray of memory cells can include a first portion of memory cells coupled to a second sensing circuitry stripe. A second portion of memory cells of the second subarray and a second portion of memory cells of the third subarray can be coupled to a third sensing circuitry stripe. A particular row of the second array can include memory cells from the first portion of memory cells in the second array coupled to memory cells from the second portion of memory cells in the second array.

Abstract: In conventional systems with a plurality of UFS devices daisy-chained to a UFS host, a UFS device driver must be able to differentiate among the links, and send either link control messages or data/management (D/M) messages to a UFS host controller. This can make force the UFS device driver to be complicated and error prone. To address this issue, a host controller can provide a uniform view of a plurality of daisy-chained devices to a device driver of a host. For example, the host controller can be such that from the perspective of the device driver, each device can appear to be a point-to-point connected device. This can allow the device driver to use a same set of link control messages to control the links. In this way, the device driver can be simplified and thus less error prone.

Abstract: A circuit includes selected sense circuits configured to be connected to selected bit lines and unselected sense circuits configured to be connected to unselected bit lines during a sense operation. The selected and unselected sense circuits are configured to perform a state-dependent pre-charge operation during the sense operation. In particular, the selected sense circuits may enable respective pre-charge circuit paths that supply a pre-charge supply voltage to respective sense nodes in the selected sense circuits. Additionally, the unselected sense circuits may disable respective pre-charge circuit paths to prevent the supply of the pre-charge supply voltage to respective sense nodes in the unselected sense circuits. A sense circuit controller may control latches to control the enabling and disabling of the pre-charge circuit paths.

Abstract: The present disclosure relates to a structure which includes at least one bit line restore device which is configured to precharge a bit line to a specified voltage during an intracycle time between a read operation and a write operation and is configured to be turned off during the read operation and the write operation.

Abstract: A structure includes a write driver circuit configured to drive both a true bitline side and a complement bitline side up to a power supply and down to ground such that one of the true bitline side and the complement bitline side is driven to ground and another of the true bitline side and the complement bitline side is driven to a high level at a same time and before a precharge below a level of the power supply of the one of the true bitline side and the complement bitline side.

Abstract: A dynamic bit-line clamping circuit for computing-in-memory applications is configured to clamp a bit line via at least one reference signal and includes a clamping node, a first clamping unit, a second clamping unit, a first feedback controlling unit and a second feedback controlling unit. The first clamping unit is electrically connected between the bit line and the clamping node. The second clamping unit is electrically connected between the clamping node and a power source voltage and includes a switch. The second feedback controlling unit is electrically connected to the clamping node and the switch. The second feedback controlling unit generates a switching signal according to the at least one reference signal and a voltage level of the clamping node. The switch is switched by the switching signal so as to clamp the voltage level of the clamping node according to the at least one reference signal.