Abstract: A verification method is disclosed. The verification method is applicable for a verification device, and the verification device is configured to verify a device under test. The device under test includes several functional blocks. The verification method includes the following operations: generating several test cases corresponding to the several functional blocks, in which the several test cases include a tree structure, in which every one of the several test cases inherits another one of the several test cases; and verifying the several functional blocks according to the several test cases and the tree structure.

Abstract: A nitride semiconductor device is provided. It includes a source electrode, a drain electrode, a gate electrode, a dielectric material layer and a stepped source field plate. The gate electrode is located between the source electrode and the drain electrode. The stepped source field plate is arranged in a part of the dielectric material layer and includes a first stepped portion and a second stepped portion respectively separated from a nitride epitaxial layer by different dielectric material layer thicknesses, and a part of the dielectric material layer between the stepped source field plate and the drain electrode is defined with a groove. By combining the stepped source field plate with the groove, parasitic capacitances Cgd and Cds are effectively reduced, the radio frequency (RF) gain of the device is improved, and the application bandwidth of the device is broadened.

Abstract: A micro inverter and a modulation method thereof are provided in the present disclosure. The micro inverter includes a filter, a transformer, and a switching circuit. The switching circuit includes one or more first switching circuit and one second switching circuit. Each of the first switching circuits has an input terminal serving as an output terminal of a direct-current power supply, and is configured to convert a direct current into an alternating current and output the alternating current to the second switching circuit. An output terminal of the second switching circuit is connected to a primary side of the transformer. A secondary side of the transformer is connected to an input terminal of the filter.

Abstract: A micro inverter, a photovoltaic system and a method for controlling a micro inverter are provided. The micro inverter includes a controller, a primary H-bridge, a transformer, a bi-directional switch arm and a capacitor arm. The controller is configured to: determine a reactive current based on a grid voltage, an equivalent capacitance of an output terminal of the inverter and a grid angle; add the reactive current to a given current to generate an internal phase-shift angle and an external phase-shift angle from a new given current obtained through the addition, where the internal phase-shift angle refers to a phase-shift angle between two arms of the primary H-bridge and the external phase-shift angle refers to a phase-shift angle between the primary H-bridge and the bi-directional switch arm.

Abstract: An end cover assembly, an air cylinder, a tread sweeper and a railway vehicle.

Abstract: An abrasive block mounting and dismounting assembly, a mounting and dismounting method, a wheel tread cleaning device, and a railway vehicle.

Abstract: This application relates to the field of image annotation and recognition in the field of artificial intelligence technologies, and in particular, to a data annotation method. The method includes: using at least two different classification models; pretraining one of the classification models as an initial classification model, and annotating a label for data in a to-be-annotated source dataset as initial data by using the pretrained classification model; and controlling the classification models to perform alternating training and data annotation a quantity of times. Operations of current training and current data annotation include: obtaining data that is re-annotated with a label by a previously trained classification model, selecting a first part of the data to train a current classification model, and re-annotating, by the trained current classification model, a label for a second part of data that is not selected.

Abstract: A model training method and a related apparatus are provided, which may be used in computer vision to perform image detection. The method includes: extracting feature information in a target image; further separately extracting features of a target object from the feature information by using a Gaussian mask to obtain a first local feature and a second local feature; determining a feature loss by using the first local feature and the second local feature; performing prediction by using the first network and the second network based on a same region proposal set to obtain a first classification predicted value and a second classification predicted value, and obtaining a classification loss based on the first classification predicted value and the second classification predicted value; and training the second network based on the classification loss and the feature loss to obtain a target network.