Abstract: The present disclosure provides methods, apparatuses and computer program products for hand-drawing shape-based document retrieval. An input hand-drawing shape may be obtained. A hand-drawing shape feature of the hand-drawing shape may be extracted through a feature extracting model. At least one target document may be retrieved by using the hand-drawing shape feature and a feature index library associated with a plurality of candidate documents, at least one document page in the target document locally matching the hand-drawing shape.

Abstract: The present disclosure relates to the technical field of electrical device state monitoring, and particularly discloses an open-phase fault monitoring system, in which a relay communication unit is configured with an MCU as a central control processing unit, and an external ADC realizes high-speed A/D conversion for a plurality of signals; a timing function is realized by a GNSS&RTC of the relay communication unit, a data interaction function is realized by a LORA module, and data is sent to a background industrial personal computer based on a designed sampling frequency. The system of the present disclosure has the advantages of small volume, high precision, low cost, easy installation and the like, and realizes real-time monitoring and accurate identification of a line open-phase fault of an auxiliary power supply system in a nuclear power plant.

Abstract: The present disclosure provides a method and apparatus for recognizing a text, a device and a storage medium, and relates to the field of deep learning technology. A specific implementation comprises: receiving a target image; performing a text detection on the target image using a pre-trained lightweight text detection network, to obtain a text detection box; and recognizing a text in the text detection box using a pre-trained lightweight text recognition network, to obtain a text recognition result.

Abstract: The present disclosure provides methods and apparatuses for multi-screen management. In some implementations, a plurality of connection requests may be received from a plurality of screen devices respectively. A screen host unit corresponding to each screen device may be created. A connection of each screen host unit to a corresponding screen device may be established. Screen display data may be transmitted, by at least one screen host unit, to at least one corresponding screen device through at least one corresponding connection respectively. Moreover, in some implementations, a connection request may be transmitted to a screen management hub. A connection to a corresponding screen host unit created by the screen management hub may be established. Screen display data may be received from the screen host unit through the connection. The screen display data may be displayed.

Abstract: The present disclosure provides a method for training a feature extraction model, a method for classifying an image and related apparatuses, and relates to the field of artificial intelligence technology such as deep learning and image recognition. The scheme comprises: extracting an image feature of each sample image in a sample image set using a basic feature extraction module of an initial feature extraction model, to obtain an initial feature vector set; performing normalization processing on each initial feature vector in the initial feature vector set using a normalization processing module of the initial feature extraction model, to obtain each normalized feature vector; and guiding training for the initial feature extraction model through a preset high discriminative loss function, to obtain a target feature extraction model as a training result.

Abstract: The present disclosure relates to the technical field of electrical device state monitoring, and particularly discloses an open-phase fault monitoring system, in which a relay communication unit is configured with an MCU as a central control processing unit, and an external ADC realizes high-speed A/D conversion for a plurality of signals; a timing function is realized by a GNSS&RTC of the relay communication unit, a data interaction function is realized by a LORA module, and data is sent to a background industrial personal computer based on a designed sampling frequency. The system of the present disclosure has the advantages of small volume, high precision, low cost, easy installation and the like, and realizes real-time monitoring and accurate identification of a line open-phase fault of an auxiliary power supply system in a nuclear power plant.

Abstract: Provided are an image processing method and apparatus, a device, a medium and a program product. The image processing method includes: performing image augmentation on an original image to obtain at least one augmented image; performing subject detection on the original image and the at least one augmented image to obtain an original detection frame in the original image and an augmented detection frame in the at least one augmented image; determining whether the original detection frame and the augmented detection frame belong to the same subject; and in response to the original detection frame and the augmented detection frame belonging to the same subject, determining a target subject frame in the original image according to the augmented detection frame.

Abstract: A system described herein can generate a graphical user interface (GUI) for a piece of forecasting software. The GUI can include graphical nodes arranged on a drag-and-drop canvas to define an overall forecasting pipeline. The graphical nodes can include a hierarchical modeling node that enables a user to define a time series hierarchy. The hierarchical modeling node can also enable separate level pipelines to be customized for each level of the time series hierarchy. The level pipelines can form subparts of the overall forecasting pipeline. The system can then execute the level pipelines to generate multiple forecasts, where each forecast corresponds to a respective level of the time series hierarchy. In some examples, the system can execute a reconciliation process on the forecasts to generate reconciled forecasts. Each reconciled forecast can correspond to one of the forecasts. The system may then generate one or more visualizations of the reconciled forecasts.

Abstract: There is provided a method and apparatus of training a model, a device, and a medium, which relate to artificial intelligence, and in particular to a deep learning and image processing technology. The method may include: determining a plurality of augmented sample sets associated with a plurality of original samples; determining a first constraint according to a first model based on the plurality of augmented sample sets; determining a second constraint according to the first model and a second model based on the plurality of augmented sample sets, wherein the second constraint is associated with a difference between outputs of the first model and the second model for one augmented sample, and the first model has a complexity lower than that of the second model; training the first model based on at least the first constraint and the second constraint, so as to obtain a trained first model.

Abstract: The present disclosure provides methods, apparatuses and computer program products for hand-drawing shape-based document retrieval. An input hand-drawing shape may be obtained. A hand-drawing shape feature of the hand-drawing shape may be extracted through a feature extracting model. At least one target document may be retrieved by using the hand-drawing shape feature and a feature index library associated with a plurality of candidate documents, at least one document page in the target document locally matching the hand-drawing shape.

Abstract: Provided are a three-phase power conversion circuit for a household electrical appliance, a household electrical appliance, and a power-on control method for a household electrical appliance. The three-phase power conversion circuit includes a three-level active PFC unit, at least two power-on units, and a control unit. The three-phase power, subsequent to being converted by the three-level active PFC unit, obtains a first DC power and a second DC power to supply power to corresponding loads. The power-on unit is arranged at input terminals of the three-level active PFC unit to buffer a device in the three-level active PFC unit when the three-level active PFC unit is powered on. A first controllable switch module and a second controllable switch module in the power-on unit and the three-level active PFC unit are controlled by the control unit to complete buffering.

Abstract: Provided are a method for controlling a full bus load to discharge in a household electrical appliance and related apparatuses. In the method, in response to determining that a half bus load is turned on and a full bus load is turned off, a voltage difference between a first DC power and a second DC power is determined to obtain a voltage difference. The first DC power and the second DC power are obtained by converting three-phase power through a three-level active PFC unit. Subsequent to the voltage difference between the first DC power and the second DC power being determined, a full bus load is controlled to be turned on or off based on the voltage difference to stabilize a half bus voltage in a predetermined voltage interval.

Abstract: The present disclosure provides methods and apparatuses for multi-screen management. In some implementations, a plurality of connection requests may be received from a plurality of screen devices respectively. A screen host unit corresponding to each screen device may be created. A connection of each screen host unit to a corresponding screen device may be established. Screen display data may be transmitted, by at least one screen host unit, to at least one corresponding screen device through at least one corresponding connection respectively. Moreover, in some implementations, a connection request may be transmitted to a screen management hub. A connection to a corresponding screen host unit created by the screen management hub may be established. Screen display data may be received from the screen host unit through the connection. The screen display data may be displayed.

Abstract: The present disclosure provides methods and apparatuses for multi-screen management. In some implementations, a plurality of connection requests may be received from a plurality of screen devices respectively. A screen host unit corresponding to each screen device may be created. A connection of each screen host unit to a corresponding screen device may be established. Screen display data may be transmitted, by at least one screen host unit, to at least one corresponding screen device through at least one corresponding connection respectively. Moreover, in some implementations, a connection request may be transmitted to a screen management hub. A connection to a corresponding screen host unit created by the screen management hub may be established. Screen display data may be received from the screen host unit through the connection. The screen display data may be displayed.

Abstract: The present disclosure provides an image recognition method and apparatus, an electronic device and a readable storage medium, and relates to the field of artificial intelligence technologies, such as image processing and deep learning technologies. The image recognition method includes: acquiring a to-be-recognized image, and determining a to-be-recognized subject in the to-be-recognized image; extracting a subject feature of the to-be-recognized subject, and obtaining a target feature according to the subject feature; determining a target candidate feature in a plurality of candidate features using the target feature; and taking a class corresponding to the target candidate feature as a recognition result of the to-be-recognized subject. With the present disclosure, different image recognition requirements may be met, and a speed and accuracy of image recognition may be improved.

Abstract: The present disclosure provides methods and apparatuses for multi-screen management. In some implementations, a plurality of connection requests may be received from a plurality of screen devices respectively. A screen host unit corresponding to each screen device may be created. A connection of each screen host unit to a corresponding screen device may be established. Screen display data may be transmitted, by at least one screen host unit, to at least one corresponding screen device through at least one corresponding connection respectively. Moreover, in some implementations, a connection request may be transmitted to a screen management hub. A connection to a corresponding screen host unit created by the screen management hub may be established. Screen display data may be received from the screen host unit through the connection. The screen display data may be displayed.

Abstract: A method for fabricating a gate-all-around (GAA) structure, including: etching a superlattice laminate to form active regions; performing selective epitaxy growth of a silicon germanium (SiGe) layer to form a SiGe-wrapped Si nanosheet stacked structure, where the SiGe layer and the SiGe/Silicon (Si) periodic superlattice laminate have the same germanium (Ge) content; after silicon oxide is backfilled and chemical mechanical polishing (CMP) is performed on the active regions, performing an amorphous-silicon dummy-gate process on a top of the active regions; removing dummy gate, and selectively etching the SiGe layer; and forming hole-trench structures connected with trenches of the dummy gate around the Si nanosheet stacked structure.

Abstract: A method for fabricating a vertical channel nanowire transistor with asymmetric stress distribution includes: (A) growing epitaxially a single-crystal material on a substrate; forming a laminate of a bottom source-drain material and a channel material; and generating a vertical uniaxial stress in the lightly-doped channel layer; (B) forming an inter-device isolation in an active layer; (C) forming a vertical channel by patterning; (D) depositing a dielectric layer to form a bottom gate isolation; (E) depositing a dummy gate layer followed by patterning to form a dummy gate pattern; (F) depositing a dielectric layer to form a top gate isolation; (G) patterning the top gate isolation; and forming a top source-drain by epitaxy growth; (H) removing a dummy gate; and forming a gate oxide layer and a metal gate; and (I) forming metal contact at individual ends of the device.

Abstract: The present disclosure provides a method for training a feature extraction model, a method for classifying an image and related apparatuses, and relates to the field of artificial intelligence technology such as deep learning and image recognition. The scheme comprises: extracting an image feature of each sample image in a sample image set using a basic feature extraction module of an initial feature extraction model, to obtain an initial feature vector set; performing normalization processing on each initial feature vector in the initial feature vector set using a normalization processing module of the initial feature extraction model, to obtain each normalized feature vector; and guiding training for the initial feature extraction model through a preset high discriminative loss function, to obtain a target feature extraction model as a training result.

Abstract: Provided are an image processing method and apparatus, a device, a medium and a program product. The image processing method includes: performing image augmentation on an original image to obtain at least one augmented image; performing subject detection on the original image and the at least one augmented image to obtain an original detection frame in the original image and an augmented detection frame in the at least one augmented image; determining whether the original detection frame and the augmented detection frame belong to the same subject; and in response to the original detection frame and the augmented detection frame belonging to the same subject, determining a target subject frame in the original image according to the augmented detection frame.