Abstract: A method of training an audio separation network is provided. The method includes obtaining a first separation sample set, the first separation sample set including at least two types of audio with dummy labels, obtaining a first sample set by performing interpolation on the first separation sample set based on perturbation data, obtaining a second separation sample set by separating the first sample set using an unsupervised network, determining losses of second separation samples in the second separation sample set, and adjusting network parameters of the unsupervised network based on the losses of the second separation samples, such that a first loss of a first separation result outputted by an adjusted unsupervised network meets a convergence condition.

Abstract: Disclosed in the present invention are a dynamic maximal clique enumeration device and method based on an FPGA with an HBM, the method including: the HBM stores a dynamic edge flow, a complete graph adjacency matrix, and candidate cliques; a matrix computing unit updates the complete graph adjacency matrix based on the dynamic edge flow, transmits the updated complete graph adjacency matrix to the HBM for storage, and determines header nodes, of which the corresponding candidate clique needs to be updated; a sequence computing unit constructs, according to the updated complete graph adjacency matrix and each header node to be updated, the sorted data set for reconstructing candidate cliques by data block sequencing; and an update computing unit executes, in parallel, an update task of the candidate clique corresponding to each header node to be updated based on the sorted data set, transmits the updated candidate cliques to the HBM for storage, and transmits the updated candidate cliques to the PC host to ext

Abstract: A method, computer program, and computer system is provided for an all-deep-learning based AEC system by recurrent neural networks. The model consists of two stages, echo estimation stage and echo suppression stage, respectively. Two different schemes for echo estimation are presented herein: linear echo estimation by multi-tap filtering on far-end reference signal and non-linear echo estimation by single-tap masking on microphone signal. A microphone signal waveform and a far-end reference signal waveform are received. An echo signal waveform is estimated based on the microphone signal waveform and a far-end reference signal waveform. A near-end speech signal waveform is output based on subtracting the estimated echo signal waveform from the microphone signal waveform, and echoes are suppressed within the near-end speech signal waveform.

Abstract: An electronic device includes an antenna unit having an antenna body having a feed point and a ground point between a first end and a second end. The antenna body has an operating band with a resonance of a first wavelength. An electrical length of the antenna body from the feed point to the ground point is greater than or equal to ¼ and less than ½ of the first wavelength. An electrical length from the first end to the feed point is greater than or equal to ? and less than or equal to ¼ of the first wavelength. An electrical length from the second end to the ground point is greater than or equal to ? and less than or equal to ¼ of the first wavelength. The antenna body is adapted to operate in a slot mode and in a wire mode.

Abstract: A method, computer program, and computer system is provided for identifying a speaker in at text based work. Labeled and unlabeled instances corresponding to one or more speakers are extracted. Pseudo-labels are inferred for the extracted unlabeled instances based on the labeled instances. One or more of the unlabeled instances are labeled based on the inferred pseudo-labels.

Abstract: A method and apparatus comprising computer code configured to cause a processor or processors to receive a text comprising a plurality of sentences, by a machine learning model, extract a nonverbal message from one of the sentences and add an annotation to the text, the annotation indicating the nonverbal message, and output a version of the text including the annotation.

Abstract: A backlight module includes a light guide plate having a bottom surface, a light exit surface, and a light incident surface a substrate located on a side of the bottom surface away from the light exit surface; first light-emitting devices located on a side of the substrate proximate to the bottom surface second light-emitting devices disposed opposite to the light incident surface; and a prism sheet located on a side of the light guide plate away from the substrate. The first light-emitting devices are configured to emit first light, so that the first light passes through the light guide plate and the prism sheet to become light of a first viewing angle. The prism sheet is configured to cooperate with the light guide plate to modulate second light emitted by the second light-emitting devices into light of a second viewing angle which is smaller than the first viewing angle.

Abstract: A method performed by at least one processor of an acoustic howling suppression (AHS) system includes receiving, from an input source device, an audio signal. The method includes refining one or more parameters of a Kalman filter based on one or more neural networks. The method includes filtering the audio signal using the Kalman filter with the one or more refined parameters of the Kalman filter to reduce acoustic howling included in the audio signal.

Abstract: A method and apparatus comprising computer code configured to cause a processor or processors to receive multiple audio signals obtained from ones of a plurality of microphones of a microphone array, implement an audio zooming based on the audio signals by selectively focusing and enhancing first ones of the audio signals and by attenuating other ones of the audio signals, and control an output of audio based on the audio zooming, and the audio zooming includes a consolidating of a plurality of directional features of the first ones of the audio signals within a field around the microphone array and a countering based on determining directional aspects of the other ones of the audio signals from outside of the field.

Abstract: A method and apparatus comprising computer code configured to cause a processor or processors to receive an audio signal obtained from a microphone, input the audio signal into a neural-network pipeline, the neural-network pipeline including a convolutional network that receives the audio signal and provides a first output of the convolutional network to an enhancer, the enhancer including a deep complex convolutional recurrent network that receives the first output along with a mel spectrogram of the audio signal and outputs a second output to at least one of a vocoder and a decoder, and control an output of an enhanced audio signal from the at least one of the vocoder and the decoder.

Abstract: An electronic device including an antenna structure is provided. A side frame between a first position and a second position of the electronic device is configured as a first radiator of the antenna structure. A side frame between the second position and a third position of the electronic device is configured as a second radiator of the antenna structure. A first slot is provided at the second position of the side frame. A second slot is formed between the second radiator and a ground. The first radiator extends in a first direction, and is spaced from the third radiator in a second direction. Projections of the third radiator and the first radiator in the second direction at least partially overlap. A first feed point is disposed on the third radiator or the first radiator, and the first feed point is configured to feed the antenna structure.

Abstract: An embodiment electric vehicle drive apparatus for decreasing a drive speed of a motor and transferring the decreased drive speed to an output shaft is provided. The apparatus includes a first planetary gear set including three first planetary gear elements in which a first element is connected to the motor and a second element is connected to a first output shaft, a second planetary gear set including three second planetary gear elements in which a first element is connected to a third element of the three first planetary gear elements, a second element is connected to a second output shaft, and a third element is fixed, and a third planetary gear set including three third planetary gear elements in which first and second elements are connected to two elements of the three first planetary gear elements, respectively, and a third element is connected to a torque vectoring motor.

Abstract: A model training method includes obtaining an image sample set and brief-prompt information; generating a content mask set according to the image sample set and the brief-prompt information; generating a to-be-trained image set according to the content mask set; obtaining, based on the image sample set and the to-be-trained image set, a predicted image set through a to-be-trained information synthesis model, the predicted image set comprising at least one predicted image, the predicted image being in correspondence to the image sample; and training, based on the predicted image set and the image sample set, the to-be-trained information synthesis model by using a target loss function, to obtain an information synthesis model.

Abstract: A method and apparatus comprising computer code configured to cause a processor or processors to receive an audio signal obtained from a microphone, input the audio signal into a neural-network based AHS model, train the neural-network based AHS model based on input signals which are recursively generated from the audio signal during training of the AHS model, and output an AHS signal from the neural-network based AHS model in which AHS is applied to the audio signal, wherein the AHS signal is a version of the audio signal in which acoustic howling noise of the audio signal is suppressed and target audio of the audio signal is sustained.

Abstract: A measurement method for the impedance of the converter at multiple operating points based on the secondary-side disturbance includes the following steps: superimposing the positive sequence current disturbance in the current sampling value; superimposing the positive sequence voltage disturbance in the voltage sampling value; determining whether measurement data of the converter at two or more current operating points are obtained, if so, based on the measurement data of the converter at two current operating points, calculating values of A(s) and B(s); further calculating values of coefficients at each frequency point, and obtaining the impedance value of the converter at multiple current operating points by calculation. The above-mentioned measurement method overcomes the disadvantages of high cost and complicated operation of the impedance measurement method based on the primary-side disturbance, and solves the problem of lacking the pre-judgment function.

Abstract: A method includes receiving, by a first encoder, an original speech segment, receiving, by a second encoder, an augmented speech segment of the original speech segment, generating, by the first encoder, a first speaker representation based on the original speech segment, generating, by the second encoder, a second speaker representation based on the augmented speech segment, and generating a contrastive loss based on the first speaker representation and the second speaker representation.

Abstract: A method, including: monitoring resource utilization of an operating system (OS) with applications utilizing larger pages; determining the monitored resource utilization is greater than a threshold resource utilization; in response to the determining the monitored resource utilization is greater than a threshold resource utilization, determining a respective larger pages index value for each of the applications utilizing larger pages; and turning off larger pages utilization of a subset of the applications utilizing larger pages, wherein the subset comprises a predefined number of the applications utilizing larger pages that have highest determined larger pages index values.

Abstract: An antenna apparatus includes a circuit board and an antenna body. The antenna body includes a first radiator and a second radiator that are indirectly coupled. The first radiator comprises a first stub and a second stub that are opposite to, but do not touch each other to form a first gap, the first stub and the second stub are located on a first side edge of the circuit board, a second gap is configured between the first stub and the first side edge, and also between the second stub and the first side edge. The second radiator is located on the circuit board to form a third gap in-between. A vertical projection of the second radiator is located on the first surface. The first stub and the second stub are electrically connected to reference ground of the circuit board separately.

Abstract: An audio processing apparatus and an audio processing method for dynamically adjusting an audio clock are provided. The audio processing apparatus includes a first interface, a buffer, a clock generator, a processor, and a second interface. The first interface receives audio data from the host. The buffer stores the audio data to generate a first audio packet and determines relationships between a data volume of the first audio packet and a first upper threshold and a first lower threshold. The second interface outputs the first audio packet and a clock signal to a codec apparatus. In response to the data volume of the first audio packet being less than the first lower threshold, the buffer outputs an underflow interrupt signal. In response to the data volume of the first audio packet being greater than the first upper threshold, the buffer outputs an overflow interrupt signal.

Abstract: The present disclosure relates to the technical field of magnesium metallurgy, and in particular to a device and method for magnesium smelting by vacuum carbothermal reduction of calcined dolomite. The device includes a reaction chamber, a condensation chamber, a first temperature regulation module and an air pressure regulation module, and the reaction chamber is communicated with the condensation chamber via a gas-guide tube. In the present disclosure, different condensation zones are utilized to sequentially condense gaseous products based on dew points, effectively preventing impurities from entering the condensation process of magnesium.