Abstract: An end-to-end differentiable image-processing pipeline for spectral Computed Tomography (CT) systems, including both Photon Counting (PCCT) and dual-energy (DECT) configurations, is provided. In the disclosed framework, spectral data acquired from a detector across two or more distinct X-ray energy spectra or energy bins are processed through a sequence of differentiable operators, including one or more differentiable preprocessing operators, a differentiable material-decomposition operator, and a differentiable image-reconstruction operator. The material-decomposition operator can be implemented using an unrolled iterative solver, an implicit-differentiation formulation, or a trained empirical proxy model. An image-domain loss between reconstructed and reference images, or other differentiable regularization criteria, is backpropagated through the pipeline to update parameters of the preprocessing operators, the material-decomposition operator, and/or a forward detector model.

Abstract: The present disclosure provides a video processing method and apparatus, a device, and a storage medium. The method includes: firstly, determining whether a type of a target video is a preset video type, where the preset video type is a video type that indicates supporting playing once for each user; determining whether a video state identifier of the target video is a hasViewed identifier if the type of the target video is determined to be the preset video type, where the video state identifier is used for characterizing whether a playing state of the target video, recorded in a server, for a current user is a hasViewed state; and then playing the target video if the video state identifier is determined not to be the hasViewed identifier.

Abstract: A data processing method and a related device are disclosed. The method includes: after a vector operator is obtained, determining whether the vector operator is capable of being converted into an equivalent matrix multiplication operator; and converting the vector operator into a corresponding target matrix multiplication operator if the vector operator is capable of being converted into the equivalent matrix multiplication operator, so that the matrix computation unit in the intelligent chip can execute the target matrix multiplication operator.

Abstract: The invention discloses methods and an electronic device for collage making. The method includes determining, at a first device, a collage template and a first set of image blocks in the collage template to receive data imported by the first device. The method further includes sending, at the first device, collage making invitation information to a second device. The collage making invitation information is associated with the collage template, at least one of the first set of image blocks, and a second set of image blocks in the collage template to receive data to imported by the second device. The method further includes obtaining, at the first device, cooperative collage data of the second device. The method further includes determining, at the first device, a collage making result. The collage making result comprises the cooperative collage data and the data imported by the first device.

Abstract: A display method for a video playing progress bar, an apparatus, a device, and a storage medium are provided. The method includes: acquiring a played duration of a target video when monitoring that a current time reaches a preset time period; determining a proportional relationship between the played duration of the target video and a total video duration; determining position information of a progress standard bar on the video playing progress bar according to the proportional relationship; determining an end position of a first color bar and a start position of a second color bar based on the position information; drawing the first color bar based on the end position of the first color bar, and drawing the second color bar based on the start position of the second color bar; and displaying the video playing progress bar of the target video based on the first and second color bar.

Abstract: A security gateway for validating ingress traffic and/or a response to the ingress traffic in a computer network, and a method for validating the same. The security gateway includes a proxy server and a gateway agent. The proxy server receives the ingress traffic and/or the response and forwards the ingress traffic and/or the response to the gateway agent. The gateway agent validates the ingress traffic and/or the response as valid ingress traffic and/or valid response or invalid ingress traffic and/or invalid response based on predetermined validation requirements. The proxy server further receives the valid ingress traffic from the gateway agent and forwards the valid ingress traffic to a backend service and/or receives the valid response from the gateway agent and forwards the valid response to a client device or blocks the invalid ingress traffic and/or invalid response.

Abstract: The present disclosure provides a Wnt family member 4/tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein zeta (Wnt4/YWHAZ) co-modified mesenchymal stem cell (MSC)-derived exosome, and a preparation method and use thereof, belonging to the technical field of skin repair. In the present disclosure, the MSC-derived exosome is modified by overexpressing adenovirus vectors expressing a Wnt4 gene and a YWHAZ gene in the MSCs, such that the exosome overexpresses the Wnt4 gene and the YWHAZ gene, to achieve a medicinal purpose of promoting regeneration and repair of a scalded skin tissue in a rat with the MSC-derived exosome.

Abstract: This application discloses a model training method and apparatus, and a device, and relates to the field of machine learning technologies. Because different computing devices have different data organization forms, a training device obtains correction data determined by each computing device based on a same training direction (a first gradient), so that the training device does not need to consider different data organization forms when training a model based on the correction data. This avoids a problem of poor stability of model training. In addition, all different computing devices run the model and output the correction data based on the same training direction. This helps the training device obtain a more accurate model training direction, thereby reducing a quantity of rounds of model training, and also reducing a quantity of times of communication between the training device and the computing devices.

Abstract: This application discloses a model training method and apparatus based on a hybrid parallelism manner. In this method, a neural network model is divided into a plurality of pipeline stages, and each pipeline stage includes a plurality of sub-stages of the neural network model. Computing nodes corresponding to the plurality of pipeline stages are invoked in a hybrid parallelism manner according to a sequence of sub-stages in the neural network model. When iterative training is performed on a network layer in a corresponding pipeline stage, because sub-stages at same locations in adjacent pipeline stages are consecutive in the neural network model, the computing node does not need to wait for completion of forward propagation of a previous pipeline stage, and can perform forward propagation on the corresponding pipeline stage only after forward propagation of the 1st sub-stage in the previous pipeline stage is completed.

Abstract: Provided are a content display method and apparatus, an electronic device, and a computer-readable storage medium. The content display method includes: displaying a first content and a first display region, in which the first display region has a first display state and a second display state, the first display state corresponding to a first state of a first object, and the second display state corresponding to a second state of the first object; and displaying, in response to detection of a trigger signal in the first display region, a second content relevant to the first object when the first display region is displayed in the first display state. A state of the first object is prompted by means of a display state of the first display region, and a content relevant to the first object is directly reached through the first display region. Thus, a problem of resource waste resulted from displaying the content relevant to the first object can be solved.

Abstract: A charging method includes: detecting that a first electronic device establishes a charging connection to a second electronic device; obtaining a charging request of the second electronic device, where the charging request includes a requested charging voltage and a requested charging current; determining that the first electronic device is in a first running state, and using the requested charging voltage and the requested charging current as a power supply voltage and a power supply current that are provided for the second electronic device; and determining that the first electronic device is in a second running state, and determining, based on the second running state and the charging request, a first power supply voltage and a first power supply current that are provided for the second electronic device.

Abstract: A desert greenhouse skeleton system and a control method thereof are provided. The desert greenhouse skeleton system includes a greenhouse skeleton including an above-ground skeleton and an underground skeleton detachably combined; freezers operatively attached to the underground skeleton, and the freezers are configured into multiple groups in an annular array manner, and at least partially placed in the sand layer; and a controller communicatively coupled to the freezers and configured to dynamically adjust the freezing parameters of the freezers based on a temperature of the sand layer and/or distribution positions of the freezers, so as to change the sand layer morphology corresponding to each freezers.

Abstract: Embodiments of this application disclose a method for detecting a backup file and a related device. The method includes: obtaining an encryption heatmap of each of a plurality of backup files; determining an encryption score of the backup file based on distribution of a target color in the encryption heatmap; constructing a sequence from the encryption score of each backup file, and performing sampling on the sequence by using a sliding window, to obtain a plurality of subsequences; and performing time sequence anomaly detection on the plurality of subsequences, and determining that a backup file corresponding to an encryption score in an abnormal subsequence is ransomware-encrypted. In this way, it can be detected, without parsing the backup file, whether the backup file is ransomware-encrypted.

Abstract: The present disclosure provides a video processing method, apparatus, and device, and a storage medium, and the method includes: in a case where it is determined that a type of a target video belongs to a preset video type, recording a first timestamp when a play method call event for the target video is monitored, the preset video type being a video type that indicates only supporting playing once for each user; recording a second timestamp when a stop callback event corresponding to the play method call event is monitored; calculating a time difference between the second timestamp and the first timestamp and determining whether the time difference is smaller than a preset time threshold; and continuing to play the target video when the time difference is determined to be smaller than the preset time threshold.

Abstract: Systems/techniques that facilitate generation of image denoising training data via photon-count splitting are provided. In various embodiments, a system can access a set of sinograms generated by a photon-counting computed tomography scanner. In various aspects, the system can split the set of sinograms into a first reduced-photon-count set of sinograms and a second reduced-photon-count set of sinograms. In various instances, the system can convert, via image reconstruction, the first reduced-photon-count set of sinograms into at least one training input image and the second reduced-photon-count set of sinograms into at least one training output image. In various cases, the system can train a deep learning neural network based on the at least one training input image and the at least one training output image.

Abstract: A desert greenhouse skeleton system and a control method thereof are provided. The desert greenhouse skeleton system includes a greenhouse skeleton including an above-ground skeleton and an underground skeleton detachably combined; freezers operatively attached to the underground skeleton, and the freezers are configured into multiple groups in an annular array manner, and at least partially placed in the sand layer; and a controller communicatively coupled to the freezers and configured to dynamically adjust the freezing parameters of the freezers based on a temperature of the sand layer and/or distribution positions of the freezers, so as to change the sand layer morphology corresponding to each freezers.

Abstract: The present invention provides use of substance A in the preparation of a medicament. The substance A is a naphthylurea compound as shown in formula I, a pharmaceutically acceptable salt thereof, a solvate thereof, a solvate of the pharmaceutically acceptable salt thereof, a crystal form thereof or a tautomer thereof, and the medicament is a medicament for treating a corneal neovascular disease. The naphthylurea compound of the present invention can treat the corneal neovascular disease by means of eye dropping administration.

Abstract: A window air conditioner includes a casing, an indoor heat exchanger, an indoor fan assembly, an outdoor heat exchanger and an outdoor fan assembly. The casing includes a base, an indoor housing and an outdoor housing; the indoor heat exchange, the indoor fan assembly, the outdoor heat exchanger and the outdoor fan assembly are arranged on the base. A top of the indoor housing is lower than a top of the outdoor housing, and an avoiding groove recessed towards the base is formed between the indoor housing and the outdoor housing.