Abstract: The present disclosure provides a motor structure including a winding set, a rotor and a motor base. The rotor is driven by the winding set. The motor base includes a central portion, a surrounding portion and at least one connecting element. The central portion is configured to dispose the winding set thereon. The surrounding portion is disposed around the central portion. The connecting element is deformable and connected between the central portion and the surrounding portion. The connecting element includes a first supporting part, a second supporting part and a deformation part. The first supporting part and the second supporting part are connected with the central portion and the surrounding portion, respectively, and the deformation part is connected therebetween. The hardnesses of the first supporting part and the second supporting part are greater than or equal to those of the central portion and the surrounding portion, respectively.

Abstract: A method, computer program, and computer system is provided for automated speech recognition. Audio data corresponding to one or more speakers is received. Covariance matrices of target speech and noise associated with the received audio data are estimated based on a gated recurrent unit-based network. A predicted target waveform corresponding to a target speaker from among the one or more speakers is generated by a minimum variance distortionless response function based on the estimated covariance matrices.

Abstract: An apparatus and a method include receiving an input audio signal to be processed by a multi-band synchronized neural vocoder. The input audio signal is separated into a plurality of frequency bands. A plurality of audio signals corresponding to the plurality of frequency bands is obtained. Each of the audio signals is downsampled, and processed by the multi-band synchronized neural vocoder. An audio output signal is generated.

Abstract: The present disclosure describes a method of fabricating a semiconductor structure that includes forming a dummy gate structure over a substrate, forming a first spacer on a sidewall of the dummy gate structure and a second spacer on the first spacer, forming a source/drain structure on the substrate, removing the second spacer, forming a dielectric structure over the source/drain structure, replacing the dummy gate structure with a metal gate structure and a capping structure on the metal gate structure, and forming an opening in the dielectric structure. The opening exposes the source/drain structure. The method further includes forming a dummy spacer on a sidewall of the opening, forming a contact structure in the opening, and removing the dummy spacer to form an air gap between the contact structure and the metal gate structure. The contact structure is in contact with the source/drain structure in the opening.

Abstract: A method for detecting a keyword, applied to a terminal, includes: extracting a speech eigenvector of a speech signal; obtaining, according to the speech eigenvector, a posterior probability of each target character being a key character in any keyword in an acquisition time period of the speech signal; obtaining confidences of at least two target character combinations according to the posterior probability of each target character; and determining that the speech signal includes the keyword upon determining that all the confidences of the at least two target character combinations meet a preset condition. The target character is a character in the speech signal whose pronunciation matches a pronunciation of the key character. Each target character combination includes at least one target character, and a confidence of a target character combination represents a probability of the target character combination being the keyword or a part of the keyword.

Abstract: A method for planning parameters of a fitness course is disclosed. The method includes the following steps: generating a plurality of limb motion signals by sensing a plurality of limb motions of a body builder through a sensing module, and sensing a physiological state of the body builder to generate a physiological state signal via the sensing module; obtaining a workout characteristic index (WCI) by performing a first calculation related to the plurality of limb motion signals, and obtaining a physiological effect index (PEI) by performing a second calculation associated with the physiological state signal; obtaining a workout effect index (WEI) by performing a third calculation associated with the WCI and the PEI; and evaluating a plurality of categorical factors associated with the WEI to plan the parameters of the fitness course.

Abstract: A fan includes a base, a motor and an impeller, and the motor includes a sleeve, a rotor, a position limiting element and a composite elastic element. The sleeve is disposed on the base, and the sleeve has a bottom portion and a side wall, which constitute an accommodation space. The rotor has a rotating shaft disposed within the accommodation space, and the rotating shaft has an outer wall. An annular groove is formed on the outer wall and has a surface. The position limiting element is disposed on the sleeve and extended from the side wall toward the rotating shaft, and the position limiting element has a bottom surface. The bottom surface faces toward the bottom portion, and the bottom surface is interfered with the surface, thereby eliminating the gap between the bottom surface and the surface. Therefore, noise of the up and down vibration is avoided from generating.

Abstract: An apparatus and a method include receiving an input audio signal to be processed by a multi-band synchronized neural vocoder. The input audio signal is separated into a plurality of frequency bands. A plurality of audio signals corresponding to the plurality of frequency bands is obtained. Each of the audio signals is downsampled, and processed by the multi-band synchronized neural vocoder. An audio output signal is generated.

Abstract: The present disclosure describes a method of fabricating a semiconductor structure that includes forming a dummy gate structure over a substrate, forming a first spacer on a sidewall of the dummy gate structure and a second spacer on the first spacer, forming a source/drain structure on the substrate, removing the second spacer, forming a dielectric structure over the source/drain structure, replacing the dummy gate structure with a metal gate structure and a capping structure on the metal gate structure, and forming an opening in the dielectric structure. The opening exposes the source/drain structure. The method further includes forming a dummy spacer on a sidewall of the opening, forming a contact structure in the opening, and removing the dummy spacer to form an air gap between the contact structure and the metal gate structure. The contact structure is in contact with the source/drain structure in the opening.

Abstract: The present disclosure provides a motor structure including a winding set, a rotor and a motor base. The rotor is driven by the winding set. The motor base includes a central portion, a surrounding portion, at least one connecting element and at least one limitation element. The winding set is disposed on the central portion. The surrounding portion is disposed around the central portion. The limitation element includes a first limitation part and a second limitation part disposed on the central portion and the surrounding portion, respectively. The connecting element is deformable and connected between the central portion and the surrounding portion. The connecting element includes a first supporting part and a second supporting part connected between a deformation part. The first supporting part, the second supporting part and the deformation part are cooperatively form a buffer zone between the first limitation part and the second limitation part.

Abstract: A method, computer system, and computer readable medium are provided for automatic speech recognition. Video data and audio data corresponding to one or more speakers is received. A minimum variance distortionless response function is applied to the received audio and video data. A predicted target waveform corresponding to a target speaker from among the one or more speakers is generated based on back-propagating the output of the applied minimum variance distortionless response function.

Abstract: A method, computer system, and computer readable medium are provided for activating speech recognition based on keyword spotting (KWS). Waveform data corresponding to one or more speakers is received. One or more direction features are extracted from the received waveform data. One or more keywords are determined from the received waveform data based on the one or more extracted features. Speech recognition is activated based on detecting the determined keyword.

Abstract: A speech recognition method, apparatus, a computer device and an electronic device for recognizing speech. The method includes receiving an audio signal obtained by a microphone array; performing a beamforming processing on the audio signal in a plurality of target directions to obtain a plurality of beam signals; performing a speech recognition on each of the plurality of beam signals to obtain a plurality of speech recognition results corresponding to the plurality of beam signals; and determining a speech recognition result of the audio signal based on the plurality of speech recognition results of the plurality of beam signals.

Abstract: A top electrode of a magnetoresistive random access memory (MRAM) device over a magnetic tunnel junction (MTJ) is formed using a film of titanium nitride oriented in a (111) crystal structure rather than a top electrode which uses tantalum, tantalum nitride, and/or a multilayer including tantalum and tantalum nitride.

Abstract: A method, computer program, and computer system for separating a target voice from among a plurality of speakers is provided. Video data associated with the plurality of speakers and audio data associated with each of the one or more speakers are received. Video feature data is extracted from the received video data. The target voice is identified from among the plurality of speakers based on the received audio data and the extracted video feature data.

Abstract: This application relates to a method of extracting an inter channel feature from a multi-channel multi-sound source mixed audio signal performed at a computing device.