Abstract: Disclosed herein is an artificial intelligence apparatus includes a memory configured to store learning target text and human speech of a person who pronounces the text, a processor configured to generate synthesized speech in which the text is pronounced by synthesized sound and extract a synthesized speech feature set including information on a feature pronounced in the synthesized speech and a human speech feature set including information on a feature pronounced in the human speech, and a learning processor configured to train a speech correction model for outputting a corrected speech feature set to allow predetermined synthesized speech to be corrected based on a human pronunciation feature when a synthesized speech feature set extracted from predetermined synthesized speech is input, based on the synthesized speech feature set and the human speech feature set.

Abstract: Embodiments of the present disclosure provide a method and an apparatus for identifying a key phrase in audio, a device and a computer readable storage medium. The method for identifying a key phrase in audio includes obtaining audio data to be identified. The method further includes identifying the key phrase in the audio data using a trained key phrase identification model. The key phrase identification model is trained based on first training data for identifying feature information of words in a first training text and second training data for identifying the key phrase in a second training text. In this way, embodiments of the present disclosure can accurately and efficiently identify key information in the audio data.

Abstract: An information processing device stores, in a keyword database, keywords extracted from speech sounds picked up by a speech-sound processing device as keywords matching keyword entries in a dictionary database of the speech-sound processing device. The information processing device receives, from the speech-sound processing device, an instruction to update the dictionary database of the speech-sound processing device, and then determines, by inference, words related to the keywords stored in the keyword database, prepares an update of the dictionary database on the basis of the keywords stored in the keyword database and the related words determined by inference, and transmits the update of the dictionary database to the speech-sound processing device.

Abstract: An object recognition method is provided. The method includes obtaining speech information of a target object in a current speech environment and position information of the target object; extracting voiceprint feature from the speech information based on a trained voiceprint matching model, to obtain voiceprint feature information; obtaining a voice confidence value corresponding to the voiceprint feature information; and obtaining an object recognition result of the target object based on the voice confidence value, the position information, and the voiceprint feature information.

Abstract: A method of updating a grammar model used during speech recognition includes obtaining a corpus including at least one word, obtaining the at least one word from the corpus, splitting the at least one obtained word into at least one segment, generating a hint for recombining the at least one segment into the at least one word, and updating the grammar model by using at least one segment comprising the hint.

Abstract: The system and method generally include identifying whether an utterance spoken by a user (e.g., customer) is a complete or incomplete sentence. For example, the system may include a partial utterance detection module that determines whether an utterance spoken by a user is a partial utterance. The detection process may include providing a detection advice code that gives a recommendation for handling the utterance of interest. If it is determined that the utterance is an incomplete sentence, then the system and method can identify the type of utterance. For example, the system may include a partial utterance classification module that predicts the class of a partial utterance. The classification process may include providing a classification advice code that gives a recommendation for handling the utterance of interest.

Abstract: Techniques for learning front-end speech recognition parameters as part of training a neural network classifier include obtaining an input speech signal, and applying front-end speech recognition parameters to extract features from the input speech signal. The extracted features may be fed through a neural network to obtain an output classification for the input speech signal, and an error measure may be computed for the output classification through comparison of the output classification with a known target classification. Back propagation may be applied to adjust one or more of the front-end parameters as one or more layers of the neural network, based on the error measure.

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 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 from a neural network having periodic indications and components of a frequency spectrum of the audio signal data inputted thereto. 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: Provided are an electronic device and a control method. The electronic device comprises: a storage unit for storing a user-based dictionary; an input unit for receiving an input sentence including a user-specific word and at least one word learned by a neural network-based language model; and a processor for determining a concept category of the user-specific word on the basis of semantic information of the input sentence, adding the user-specific word to the user-based dictionary to perform update, and when text corresponding to semantic information of the at least one learned word is input, providing the user-specific word as an autocomplete recommendation word which can be input subsequent to the text.

Abstract: Provided are a device and a method for providing a response message to a voice input of a user. The method, performed by a device, of providing a response message to a voice input of a user includes: receiving the voice input of the user; determining a destination of the user and an intention of the user, by analyzing the received voice input; obtaining association information related to the destination; generating the response message that recommends a substitute destination related to the intention of the user, based on the obtained association information; and displaying the generated response message.

Abstract: Techniques for determining a command or intent likely to be subsequently invoked by a user of a system are described. A user inputs a command (either via a spoken utterance or textual input) to a system. The system determines content responsive to the command. The system also determines a second command or corresponding intent likely to be invoked by the user subsequent to the previous command. Such determination may involve analyzing pairs of intents, with each pair being associated with a probability that one intent of the pair will be invoked by a user subsequent to a second intent of the pair. The system then outputs first content responsive to the first command and second content soliciting the user as to whether the system to execute the second command.

Abstract: The present invention discloses a speech interaction method and apparatus, and pertains to the field of speech processing technologies. The method includes: acquiring speech data of a user; performing user attribute recognition on the speech data to obtain a first user attribute recognition result; performing content recognition on the speech data to obtain a content recognition result of the speech data; and performing a corresponding operation according to at least the first user attribute recognition result and the content recognition result, so as to respond to the speech data. According to the present invention, after speech data is acquired, user attribute recognition and content recognition are separately performed on the speech data to obtain a first user attribute recognition result and a content recognition result, and a corresponding operation is performed according to at least the first user attribute recognition result and the content recognition result.

Abstract: A lip-language identification method and an apparatus thereof, an augmented reality device and a storage medium. The lip-language identification method includes: acquiring a sequence of face images for an object to be identified; performing lip-language identification based on a sequence of face images so as to determine semantic information of speech content of the object to be identified corresponding to lip actions in a face image; and outputting the semantic information.

Abstract: Approaches presented herein enable communication of audio messages using augmented reality. More specifically, a set of captured messages are analyzed to determine, for each message of the set of messages, whether the message is relevant to a user. At least one message of the set of messages is associated with a real-world object over which to display the at least one message in an augmented reality environment. A time limit is assigned to the at least one message based on the analyzed relevance to the user. In the case that a current time is within the time limit, the at least one message is displayed on a display of a mobile device associated with the user when a camera of the mobile device is directed at the real-world object associated with the at least one message.

Abstract: A voice to text model used by a voice-enabled electronic device is dynamically and in a context-sensitive manner updated to facilitate recognition of entities that potentially may be spoken by a user in a voice input directed to the voice-enabled electronic device. The dynamic update to the voice to text model may be performed, for example, based upon processing of a first portion of a voice input, e.g., based upon detection of a particular type of voice action, and may be targeted to facilitate the recognition of entities that may occur in a later portion of the same voice input, e.g., entities that are particularly relevant to one or more parameters associated with a detected type of voice action.

Abstract: A process for dispatcher monitoring of an electronic digital assistant (EDA) in a group-based mobile radio system includes first receiving, at a radio controller from a mobile radio, a request to establish an EDA call for interacting with an EDA. Identifying one of (i) a primary talkgroup to which the mobile radio is affiliated and (ii) an agency to which the mobile radio belongs. Selecting or creating an EDA talkgroup, different from the primary talkgroup, through which the mobile radio may interact with the EDA. Identifying one or more, but less than all, dispatch consoles that are (i) subscribed to monitor the primary talkgroup of the mobile radio or (ii) a member of a same agency as the mobile radio. Adding the EDA, mobile radio, and identified dispatch consoles to the EDA talkgroup. And causing the EDA call to be established.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for hotword detection on multiple devices are disclosed. In one aspect, a method includes the actions of receiving, by a first computing device, audio data that corresponds to an utterance. The actions further include determining a first value corresponding to a likelihood that the utterance includes a hotword. The actions further include receiving a second value corresponding to a likelihood that the utterance includes the hotword, the second value being determined by a second computing device. The actions further include comparing the first value and the second value. The actions further include based on comparing the first value to the second value, initiating speech recognition processing on the audio data.

Abstract: Methods and systems for classifying a multimedia file using interclass data is disclosed. One of the methods includes receiving, from a first transcription engine, one or more transcription results of one or more audio segments of the multimedia file; identifying a first transcription result for a first audio segment having a low confidence of accuracy; identifying a first image data of the multimedia file corresponding to the first segment; receiving, from an image classification engine trained to classify image data, an image classification result of one or more portions of the first image data in response to requesting the image classification engine to classify the first image data; and selecting, based at least on the image classification result of the one or more portions of the first image data, a second transcription engine to re-classify the first audio segment.

Abstract: An electronic device for performing speech recognition and a method therefor are provided. The method includes detecting a first text, which is preset for performing speaker recognition, by performing speech recognition on a first speech signal, performing speaker recognition on a second speech signal acquired after the first speech signal, based on the first text being detected, and executing a voice command obtained from the second speech signal, based on a result of performing the speaker recognition on the second speech signal indicating that a speaker of the second speech signal corresponds to a first speaker who registered the first text.

Abstract: An electronic device according to various embodiments of the present invention includes: a microphone; a communication module; a memory; and at least one processor, wherein the processor can receive and record a voice through the microphone while a function of receiving the voice is activated, generate first authentication data including data for the voice and identification data for the electronic device on the basis of the recorded voice, determine the mode of the electronic device on the basis of the recorded voice, send the first authentication data, receive second authentication data corresponding to the first authentication data, use identification data included in the second authentication data to connect communication with an external electronic device when the data for the voice included in the first authentication data matches data for voice included in the second authentication data, and perform, according to the mode, at least one function related to the communication-connected external electronic

Abstract: A method to encode audio signals is provided for use with an audio capture device that includes multiple microphones having a spatial arrangement on the device, a method to encode audio signals comprising: receiving multiple microphone signals corresponding to the multiple microphones; determining a number and directions of arrival of directional audio sources represented in the one or more microphone signals; determining one of an active microphone signal component and a passive microphone signal component, based upon the determined number and directions of arrival; determining the other of the active microphone signal component and the passive microphone signal component, based upon the determined one of the active input spatial audio signal component and the passive input spatial audio signal component; encoding the active microphone signal component; encoding the passive microphone signal component.

Abstract: A method to decode audio signals is provided that includes: receiving an input spatial audio signal; determining directions of arrival of directional audio sources represented in the received input spatial audio signal; determining one of an active input spatial audio signal component and a passive spatial audio signal input component, based upon the determined directions of arrival; determining the other of the active input spatial audio signal component and the passive input spatial audio signal component, based upon the determined one of the active input spatial audio signal component and the passive input spatial audio signal component; decoding the active input spatial audio signal component to a first output format; and decoding the passive input spatial audio signal component to a second output format.

Abstract: The present invention relates to an ambisonic encoder for a sound wave having a plurality of reflections. The ambisonic encoder according to the invention makes it possible to improve the sensation of immersion in a 3D audio scene. The complexity of encoding of the reflections of sound sources for an ambisonic encoder according to the invention is less than the complexity of encoding of the reflections of sound sources of an ambisonic encoder according to the prior art. The ambisonic encoder according to the invention makes it possible to encode a greater number of reflections of a sound source in real time. The ambisonic encoder according to the invention makes it possible to reduce the power consumption related to ambisonic encoding, and to increase the life of a battery of a mobile device used for said application.

Abstract: A decoder for generating an audio output signal having one or more audio output channels is provided. The decoder includes a receiving interface for receiving an audio input signal including a plurality of audio object signals, for receiving loudness information on the audio object signals, and for receiving rendering information indicating whether one or more of the audio object signals shall be amplified or attenuated. Moreover, the decoder includes a signal processor for generating the one or more audio output channels of the audio output signal. The signal processor is configured to determine a loudness compensation value depending on the loudness information and depending on the rendering information. Furthermore, the signal processor is configured to generate the one or more audio output channels of the audio output signal from the audio input signal depending on the rendering information and depending on the loudness compensation value. Moreover, an encoder is provided.

Abstract: Systems and methods for audio signal processing including an input interface to receive a noisy audio signal including a mixture of target audio signal and noise. An encoder to map each time-frequency bin of the noisy audio signal to one or more phase-related value from one or more phase quantization codebook of phase-related values indicative of the phase of the target signal. Calculate, for each time-frequency bin of the noisy audio signal, a magnitude ratio value indicative of a ratio of a magnitude of the target audio signal to a magnitude of the noisy audio signal. A filter to cancel the noise from the noisy audio signal based on the phase-related values and the magnitude ratio values to produce an enhanced audio signal. An output interface to output the enhanced audio signal.

Abstract: Cardioid adaptive filtering includes receiving a first and second audio signals from first and second omnidirectional microphones; combining the audio input signals into a cardioid signal; filtering the cardioid signal to create a first filtered output using an adaptive low pass filter controlled by a frequency control, the adaptive low pass filter having a controllable corner frequency f1; filtering the first filtered output, using a high frequency gain filter with a corner frequency f2, to create an equalized cardioid output signal; performing feedforward processing of the audio input signals to provide a wind feedforward signal; using the equalized cardioid output and the first or second audio input signal, performing proximity feedback to generate a proximity feedback signal; adjusting the frequency f1 of the adaptive low pass filter using the wind feedforward signal and the proximity feedback signal; and providing the equalized cardioid output signal for use in receiving captured audio.

Abstract: A method of processing a signal includes taking a signal recorded by a plurality of signal recorders, applying at least one super-resolution technique to the signal to produce an oscillator peak representation of the signal comprising a plurality of frequency components for a plurality of oscillator peaks, computing at least one Cross Channel Complex Spectral Phase Evolution (XCSPE) attribute for the signal to produce a measure of a spatial evolution of the plurality of oscillator peaks between the signal, identifying a known predicted XCSPE curve (PXC) trace corresponding to the frequency components and at least one XCSPE attribute of the plurality of oscillator peaks and utilizing the identified PXC trace to determine a spatial attribute corresponding to an origin of the signal.

Abstract: Audio systems and methods are provided that enhance a portion of audio content relative to other portions of the audio content. The systems and methods select the portion to be enhanced and calculate an intelligibility metric of the selected portion, such as a dialogue portion. The systems and methods determine a gain based at least in part upon the intelligibility metric and apply the gain to the selected portion to provide an enhanced portion. The systems and methods provide an audio signal, based at least in part upon the enhanced portion, to an output for conversion to an acoustic signal, such as by an acoustic transducer.

Abstract: Disclosed herein are systems, methods, and computer-readable storage media for improving speech recognition accuracy using textual context. The method includes retrieving a recorded utterance, capturing text from a device display associated with the spoken dialog and viewed by one party to the recorded utterance, and identifying words in the captured text that are relevant to the recorded utterance. The method further includes adding the identified words to a dynamic language model, and recognizing the recorded utterance using the dynamic language model. The recorded utterance can be a spoken dialog. A time stamp can be assigned to each identified word. The method can include adding identified words to and/or removing identified words from the dynamic language model based on their respective time stamps. A screen scraper can capture text from the device display associated with the recorded utterance. The device display can contain customer service data.

Abstract: A response is prevented from being made by a malfunction. A control section (10) includes: a speech sound obtaining section (11) configured to distinctively obtain detected sounds from respective microphones (30), the detected sounds being ones that have been detected by the respective microphones (30); a noise determining section (14) configured to determine whether or not each of the detected sounds is a noise and configured to, in a case where a content of a speech is not recognized from a detected sound, determine that the detected sound is a noise; and a detection control section (17) configured to, in a case where the noise determining section (14) determines that any of the detected sounds is a noise, control at least one of the microphones (30) to stop detecting a sound.

Abstract: This application discloses a voice activity detection method. The method includes receiving speech data, the speech data including a multi-frame speech signal; determining energy and spectral entropy of a frame of speech signal; calculating a square root of the energy of the speech signal and/or calculating a square root of the spectral entropy of the frame of the speech signal; determining a spectral entropy-energy square root of the frame of the speech signal based on at least one of the square root of the energy and the square root of the spectral entropy; and determining that the frame of the speech signal is an unvoiced frame if the spectral entropy-energy square root of the speech signal is less than a first threshold, or that it is a voiced frame if the spectral entropy-energy square root of the speech signal is greater than or equal to the first threshold.

Abstract: A tonearm unit includes: an arm member having a tip to which a cartridge is attached; and a raising and lowering mechanism which raises and lowers the arm member with respect to a phonograph record along an axis. The raising and lowering mechanism includes: a cam base supporting the arm member; a ring cam provided to be rotatable about the axis with respect to the cam base; a first guide portion on the cam base, extending at an angle along a rotation direction of the ring cam; and a second guide portion on the ring cam, extending at an angle along the rotation direction so as to face the first guide portion. When the ring cam rotates with respect to the cam base, the cam base is raised and lowered with respect to the ring cam as a result of the second guide portion sliding along the first guide portion.

Abstract: According to one embodiment, a magnetic head includes a magnetic pole, a first shield, and a stacked body. The stacked body is provided between the magnetic pole and the first shield. The stacked body includes a magnetic layer including at least one selected from the group consisting of Fe, Co, and Ni, a first conductive layer provided between the magnetic pole and the magnetic layer, the first conductive layer being nonmagnetic, and a second conductive layer provided between the magnetic layer and the first shield, the second conductive layer being nonmagnetic. The first conductive layer includes Ir. A thickness of the first conductive layer along a first direction is not less than 0.3 nm and not more than 0.8 nm. The first direction is from the first conductive layer toward the second conductive layer.

Abstract: A method includes moving a heat-assisted magnetic recording (HAMR) slider relative to a magnetic recording medium. The slider comprises a writer, a writer heater, and a near-field transducer (NFT). For each of a plurality of different head-to media spacings a test tone is written to a track of the medium, the test tone is read and a Discrete Fourier Transform (DFT) of an amplitude of the read test tone is captured. A first DFT curve is generated at a beginning of writing the test tones. A second DFT curve is generated at a saturated state of writing the test tones. An amount of horizontal shift between the first and second DFT curves is computed. The amount of horizontal shift corresponding to writer heater power required to compensate for NFT clearance offset due to laser induced writer protrusion.

Abstract: A method of operating a magnetic tape within a tape drive. The tape drive comprises a tape head, which shows a tape-bearing surface meant to face a front side of a magnetic tape, in operation. The tape-bearing surface comprises a transducer area. This area includes at least one transducer, which is a read or write element configured to read or write to the magnetic tape, respectively. The method may include driving the tape (along a longitudinal direction of circulation thereof above the tape-bearing surface) and concomitantly ejecting a gas flow toward the transducer area. The gas flow ejected impinges on the back side of the driven tape (e.g., opposite to the front side of the tape), so as to locally urge the front side of the tape against the transducer area and thereby read or write to the tape via said at least one transducer.

Abstract: A system, according to one embodiment, includes: a module having transducers positioned along a tape bearing surface of the module, and a roller guide. A patterned bar is also positioned relative to the roller guide and module to engage a magnetic tape. Moreover, the patterned bar has a plurality of recessed regions along a tape bearing surface thereof. The patterned bar is also not directly coupled to the module, and both edges of the patterned bar which engage the magnetic tape are skiving edges. Other systems, methods, and computer program products are described in additional embodiments.

Abstract: A polymorphic playback system is disclosed in which one or more parameters of a signal path of the polymorphic playback system are varied based on one or more characteristics of a playback signal processed by the signal path, wherein the polymorphic playback system may include a lower-latency detection filter, a higher-latency detection filter, and a control subsystem that uses the lower-latency detection filter for detecting the one or more first characteristics of the playback signal and uses the higher-latency detection filter for detecting the one or more second characteristics of the playback signal.

Abstract: Devices and techniques for NAND cell encoding to improve data integrity are disclosed herein. A high-temperature indicator is obtained and a write operation is received. The write operation is then performed on a NAND cell using a modified encoding in response to the high-temperature indicator. The modified encoding includes a reduced number of voltage distribution positions from an unmodified encoding without changing voltage distribution widths, where each voltage distribution corresponds to a discrete set of states an encoding.

Abstract: A memory circuit array includes a first read device and a first program device. The first read device is coupled to a first bit line. The first read device includes a first transistor coupled to a first word line, and a second transistor coupled to the first word line. The first program device is coupled to the first read device. The first program device includes a third transistor coupled to a second word line, and a fourth transistor coupled to the second word line.

Abstract: Apparatuses for providing a clock signal for a plurality of circuits of a semiconductor device within delays in a certain range are described. An example apparatus includes a signal wire including a first portion and a second portion, having one ends coupled to each other at a signal input and the other ends coupled to each other that extend in parallel. The second portion has a higher impedance than the first portion from the first end to the second end. Output buffers closer to the signal input are coupled to the second portion and output buffers farther to the signal input are coupled to the first portion.

Abstract: Systems and methods are disclosed for providing micro-idle memory power management. One embodiment of a method comprises receiving and storing an exit latency vote from each of a plurality of memory subsystems on a system on chip electrically coupled to a system memory. In response to a micro-idle memory state in which each of the memory subsystems are idle, a minimum exit latency value from the plurality of exit latency votes is determined. One of a plurality of system memory modes is selected which has a micro-idle sleep time that meets the minimum exit latency value while minimizing system memory power consumption. The selected system memory mode is initiated.

Abstract: Disclosed is an improved load switch driver for Power Management Integrated Circuit (PMIC) devices. In one embodiment, a PMIC is disclosed comprising a gate driver, the gate driver connected to the gate of a switch; an operation frequency generator connected to the gate driver and configured to supply a periodic voltage to the gate driver; and a voltage sensor, the voltage sensor connected to the operation frequency generator and the source of the switch, the voltage sensor configured to monitor a drain-source voltage of the switch and lower the frequency of the operation frequency generator to a second frequency in response to detecting a collapse of the drain-source voltage.

Abstract: A calibration circuit includes first and second pull-up units each receiving a pull-up code and connected between a pad connected with an external resistor and a first power supply voltage, a pull-down unit connected between the pad and a second power supply voltage and receiving a pull-down code, a comparator comparing a first voltage with a reference voltage and then compare a second voltage with the reference voltage, a first digital filter adjusting the pull-up code based on a first comparison result of the first voltage with the reference voltage, and a second digital filter adjusting the pull-down code based on a second comparison result of the second voltage with the reference voltage.

Abstract: A write-in date circuit in a semiconductor device may include multiple input buffers, each receiving multiple data bits in a serial data stream. The circuit may include a first circuit coupled to a first and a second input buffers. The first circuit may be further coupled to receive a DQS signal and latch a first data bit selected from the first input buffer or the second input buffer responsive to the DQS signal. The second circuit may be coupled to the first and second input buffers and configured to latch a second data bit selected from the first input buffer or the second input buffer responsive to the DQS signal. The first circuit may latch the first data bit responsive to a rising edge of the DQS signal and the second circuit may latch the second data bit responsive to a falling edge of the DQS signal.

Abstract: Aspects of the present disclosure include a memory sub-system configured to reduce latency and power consumption during a read-write cycle. The memory system comprises a first memory component and a processing device operatively coupled to the first memory component. The processing device is configured to receive a request to write a first sequence of data bits from a first data block of a second memory component to memory media of the first memory component. In response to receiving the request, the processing device reads a second sequence of data bits from a second data block stored in the memory media of the first memory component, and compares the first sequence of data bits with the second sequence of data bits.

Abstract: A delay tracking method and a memory system are provided. The delay tracking method is applied to a memory system supporting a low-frequency-mode (LFM) and a high-frequency-mode (HFM) of an operating clock. The delay tracking method includes the steps of selecting a LFM oscillator for obtaining a LFM delay value when the operating clock is in the HFM; and selecting a HFM oscillator for obtaining a HFM delay value when the operating clock is in the LFM.

Abstract: A method includes performing a first write leveling training operation and performing a second write leveling training operation. The first write leveling training operation is performed to generate transmission data based on a data strobe signal and an internal command pulse and to generate a latency code. The second write leveling training operation is performed to generate the transmission data based on the data strobe signal and the internal command pulse.

Abstract: A semiconductor device may include a power supply circuit, a word line control circuit, and a memory circuit. The power supply circuit may drive a pre-charge voltage to a level of an external voltage based on a write initialization signal which is enabled if a command has a predetermined level combination. The word line control circuit may generates two or more word line selection signals that are sequentially counted based on the write initialization signal. The memory circuit may sequentially select a plurality of word lines based on the word line selection signals. The memory circuit may drive bit lines of memory cells connected to the selected word line to the pre-charge voltage. The memory circuit may store data, which are loaded on the bit lines to have a level of the pre-charge voltage, into the memory cells connected to the selected word line.

Abstract: According to one embodiment, a magnetic memory device includes a conductive member, a first magnetic layer, a second magnetic layer, and a first nonmagnetic layer. The conductive member includes a first layer. The first layer includes at least one selected from the group consisting of HfN having a NaCl structure, HfN having a fcc structure, and HfC having a NaCl structure. The first magnetic layer is separated from the first layer in a first direction. The second magnetic layer is provided between the first layer and the first magnetic layer. The first nonmagnetic layer is provided between the first magnetic layer and the second magnetic layer.

Abstract: A structure and operation method of a spin device using a magnetic domain wall movement by spin orbit torque are provided. It is possible to invert the magnetization of free layer of the device at a low value of current by using the spin orbital torque, and the structure of the device is simpler than that of the conventional CMOS. Further, a spin synapse device to which a free layer of multiaxial anisotropy is applied in addition to movement of a magnetic domain wall is provided. Since the magnetoresistance can be adjusted according to the angle of the pinned layer and the free layer, it is easy to apply multi-bit and it can be applied to artificial synapse technology.