Abstract: This application discloses a calibration method for navigation of an unmanned aerial vehicle (UAV), a non-transitory computer-readable storage medium and a UAV implementing the same. The calibration method includes: collecting, during a flight of the UAV, reference data during two measurements of a reference vector performed by a vector sensor; acquiring a zero-point offset M0 of the vector sensor according to the reference data; acquiring original data Rk of any vector measured by the vector sensor; acquiring valid data Vk according to the zero-point offset M0 and the original data Rk; and control headings and postures of the UAV according to the valid data Vk. With the calibration method in this application, the valid data Vk is defined as a vector data acquired after a zero-point error of the original data Rk is eliminated, which is more closely approximated to an actual value of a to-be-measured vector.

Abstract: This application discloses a calibration method for navigation of an UAV including a vector sensor. The calibration method includes: collecting, during a flight of the UAV, a current correction value and current data during a current measurement performed by the vector sensor; acquiring previous data during a previous measurement performed by the vector sensor; acquiring an adjustment quantity according to the current data and the previous data; acquiring a next correction value according to the current correction value and the adjustment quantity; and acquiring next original data during a next measurement performed by the vector sensor, acquiring next valid data according to the next original data and the next correction value , and controlling headings and postures of the UAV according to the next valid data. With the calibration method of this application, the next valid data Vk+1 more closely approximated to a true value can be obtained.

Abstract: This application discloses a vector sensor zero-point calibration method, device, and, apparatus, an electronic device, and a non-volatile computer-readable storage medium. The calibration method includes: acquiring reference data during two measurements of a reference vector performed by a vector sensor; acquiring a zero-point offset M0 of the vector sensor according to the reference data; acquiring original data Rk of any vector measured by the vector sensor; and acquiring valid data Vk according to the zero-point offset M0 and the original data Rk. With the calibration method in this application, the valid data Vk is obtained after a zero-point error of the original data Rk is eliminated, which is more closely approximated to an actual value of a to-be-measured vector.

Abstract: Disclosed is a nursing machine that includes a wearable working head and a urine detection circuit. The urine detection circuit includes at least two first urine sensors and at least two second urine sensors that are arranged on the working head. In response to more than one first urine sensor each outputting a liquid detected signal or in response to more than one second urine sensor each outputting the liquid detected signal, the urine detection circuit is configured to output a urine signal. In response to only one first urine sensor or only one second urine sensor outputting the liquid detected signal or in response to only one first urine sensor and only one second urine sensor each outputting the liquid detected signal, the urine detection circuit doesn't output the urine signal.

Abstract: A laser ranging device and a robot. The laser ranging device comprises an emitting unit and a receiving unit. The emitting unit comprises an emitter configured to emit a laser pulse to a target object that is to be ranged, and an emitting lens configured for the laser pulse to pass through. The receiving unit comprises a photodetector configured to receive a reflected laser pulse by the target object, and a receiving lens configured for the reflected laser pulse to pass through. The photodetector is encapsulated in an encapsulation module, the encapsulation module is disposed aft of and faces the receiving lens, and a side surface of the encapsulation module facing the receiving lens is constructed as an extinction surface.

Abstract: Computing systems and technical methods that transform data structures and pierce opacity difficulties associated with complex machine learning modules are disclosed. Advances include a framework and techniques that include: i) global diagnostics; ii) locally interpretable models LIME-SUP-R and LIME-SUP-D; and iii) explainable neural networks. Advances also include integrating LIME-SUP-R and LIME-SUP-D approaches that create a transformed data structure and replicated modeling over local and global effects and that yield high interpretability along with high accuracy of the replicated complex machine learning modules that make up a machine learning application.

Abstract: This application provides a video information generation method, apparatus, and system and a storage medium. The video information generation method includes: obtaining a plurality of temporally consecutive target images; obtaining first information of a target object in the target images; and associating first information of a same target object located in different target images to generate target information. In the video information generation method provided in this application, the first information of the target object in the target images is obtained, and the first information of the same target object located in different target images is associated. In this way, target information with a relatively small amount of data can be obtained, thereby improving the efficiency of remotely viewing a video by a user.

Abstract: A speech separation server comprises a deep-learning encoder with nonlinear activation. The encoder is programmed to take a mixture audio waveform in the time domain, learn generalized patterns from the mixture audio waveform, and generate an encoded representation that effectively characterizes the mixture audio waveform for speech separation.

Abstract: The disclosure relates to bacillus having an effect of inducing corn to resist Northern leaf blight and applications of the bacillus. The bacillus having an effect of inducing corn to resist Northern corn leaf blight is Bacillus megaterium which was deposited on Apr. 16, 2019 with a deposit number of CGMCC No. 17573. The Bacillus megaterium of the disclosure can induce corn to generate resistance to Northern corn leaf blight during seed germination and corn growth to activate the immune response of corns and stimulate immune control of Northern leaf blight of corn. The method described herein is long in coming does not pollute environment and soil, is a green, simplified and new control method of Northern leaf blight of corn, and has wide application and popularization prospect.

Abstract: A range-finding method, a apparatus, a terminal system, and a computer-readable storage medium are provided.

Abstract: A domain adaptation module is used to optimize a first domain derived from a second domain using respective outputs from respective parallel hidden layers of the domains.

Abstract: Novel methods and systems for adapting a voice cloning synthesizer for a new speaker using real speech data are disclosed. Utterances from one or more target speakers are parameterized and are used to initialize an embedding vector for use with a voice synthesizer, by means of clustering the utterance data and determining the centroid of the data, using a speaker identification neural network, and/or by finding the closest stored embedded vector to the utterance data.

Abstract: A monocular visual simultaneous localization and mapping (SLAM) data processing method.

Abstract: Automatically recommending sound effects based on visual scenes enables sound engineers during video production of computer simulations, such as movies and video games. This recommendation engine may be accomplished by classifying SFX and using a machine learning engine to output a first of the classified SFX for a first computer simulation based on learned correlations between video attributes of the first computer simulation and the classified SFX.

Abstract: The present invention discloses an improved model-free adaptive control method, in particular to an improved method for a compact dynamic linearization model-free adaptive control based on MIMO systems, and belongs to the field of control algorithm design. Firstly, proportional control is added in CFDL-MFAC to improve the problems of low response speed and large overshoot in the original control system. Secondly, an anti-windup control algorithm of an actuator is added in the above control structure so that the actuator does not conduct transfinite operation when reaching the upper or lower saturation limit and the actuator can quickly make a control response when a control instruction enters an unsaturated region again to improve the control accuracy of the system. Then, it is proved through strict analysis that the improved control algorithm can ensure tracking error and BIBO stability under certain conditions.

Abstract: Embodiments of this disclosure provide a speech enhancement method and apparatus, an electronic device, and a computer-readable storage medium. The method includes: obtaining a clean speech sample; decomposing the clean speech sample to obtain a first sparse matrix and m base matrices, values in the first sparse matrix being all positive numbers, and m being a positive integer greater than 1; obtaining, according to the first sparse matrix and a weight matrix of a target neural network, state vectors of neurons in a visible layer of the target neural network; and updating the weight matrix according to the state vectors of the neurons in the visible layer and the clean speech sample, to obtain a deep dictionary used for speech enhancement.

Abstract: The present invention discloses a metal gate structure, comprising from inside to outside: a metal electrode layer, a bonding layer surrounding the metal electrode layer, a first diffusion barrier layer and a work function layer, a second diffusion barrier layer within the first diffusion barrier layer; wherein, the second diffusion barrier layer is a composite barrier layer structure comprising a metal layer and a metal oxide layer within the metal layer.

Abstract: Provided is a vacuum cleaner. The switching valve apparatus of the vacuum cleaner includes an outer casing and a counterflow air passage. The outer casing is provided with a first vent opening and a second vent opening arranged side by side along the circumferential direction of the outer casing. In the case where the counterflow air passage is at the first angular position, the outer casing seals the counterflow air outlet of the counterflow air passage; in the case where the counterflow air passage is at the second angular position, the counterflow air outlet communicates with the first vent opening and is staggered with the second vent opening; and in the case where the counterflow air passage is at the third angular position, the counterflow air outlet communicates with the second vent opening and is staggered with the first vent opening.

Abstract: An industrial 5G dynamic multi-priority multi-access method based on deep reinforcement learning includes the following steps: establishing an industrial 5G network model; establishing a dynamic multi-priority multi-channel access neural network model based on deep reinforcement learning; collecting state, action and reward information of multiple time slots of all industrial 5G terminals in the industrial 5G network as training data; training the neural network model by using the collected data until the packet loss ratio and end-to-end latency meet industrial communication requirements; collecting the state information of all the industrial 5G terminals in the industrial 5G network at the current time slot as the input of the neural network model; conducting multi-priority channel allocation; and conducting multi-access by the industrial 5G terminals according to a channel allocation result.

Abstract: A method for preparing a NdFeB permanent magnetic material may include providing a covered NdFeB magnetic powder by depositing heavy rare earth particles or high-melting particles onto a NdFeB magnetic powder by physical vapor deposition; and performing orientation molding and sintering on the covered NdFeB magnetic powder to provide the NdFeB permanent magnetic material.