Abstract: A network security system that classifies webpages and uses the classification of the webpages to enforce relevant security policies is disclosed. To classify a webpage, a screenshot of at least a portion of the webpage is captured. An embedding engine generates a subject image embedding of the screenshot. The subject image embedding is classified by an image classifier that includes an index of training image embeddings each having a label of their classification and an associated approximate nearest neighbors model trained to identify the label of the closest training image embedding to the subject image embedding and a score representing their similarity. The webpage is classified based at least in part on the score and the label from the image classifier. The network security system applies security policies to requests from client devices that identify the webpage as its destination based at least in part on the webpage classification.

Abstract: A mobile liquid nitrogen fire extinguishing system for a utility tunnel is provided, including a liquid nitrogen production assembly, a fire extinguishing assembly, a monitoring device, and a liquid nitrogen conveying tubing. The mobile fire extinguishing assembly includes a liquid nitrogen storage tank and a liquid nitrogen booster pump. The liquid nitrogen storage tank is connected to the liquid nitrogen booster pump. The liquid nitrogen booster pump conveys liquid nitrogen to branch pipes in the utility tunnel through the liquid nitrogen conveying tubing. Liquid nitrogen injection ports are formed in each of the branch pipes. The liquid nitrogen production assembly is mobile and conveys the prepared liquid nitrogen to the liquid nitrogen storage tank. The monitoring device includes a first monitoring assembly and a second monitoring assembly.

Abstract: Disclosed is a method of building a customized deep learning (DL) stack classifier to detect organization sensitive data in images, referred to as image-borne organization sensitive documents, and protecting against loss of the image-borne organization sensitive documents, including distributing a trained feature map extractor stack, with stored parameters, configured to allow the organization to extract from image-borne organization sensitive documents, feature maps that are used to generate updated DL stacks and to save non invertible feature maps derived from the images, and ground truth labels for the image. Also included is receiving organization-specific examples including the non-invertible feature maps extracted from the organization-sensitive documents and the ground truth labels and using the received organization-specific examples to update a customer-specific DL stack classifier. Further included is sending the customer-specific DL stack classifier to the organization.

Abstract: A method for estimating the state of health of lithium-ion batteries considering user charging behavior is provided. Firstly, data during the cycle charging process of the actual vehicles are collected, and it is analyzed. Then, the charging process data is applied to extract health features. The obtained health features are used to establish a state of health estimation model, and the established model is used to estimate the state of health of the battery. It involves reading the charging data, analyzing the frequency and occurrence of the start charging voltage and stop charging voltage, integrating them into a heat map, and drawing the IC curve of the charging and discharging process. This enables any charging behavior to extract the health features of battery aging, and predicts the health values of the battery, thus completing the estimation of the state of health for lithium-ion batteries.

Abstract: A power conversion device includes a DC-to-DC converter and a control unit, the control unit includes a phase-locked loop controller, and the DC-to-DC converter includes a primary side circuit, a secondary side circuit and a transformer. The transformer transfers energy between the primary side circuit and the secondary side circuit, the primary side circuit includes an inverter circuit, and the phase-locked loop controller outputs a first driving signal. The first driving signal is generated according to a phase jitter signal, a phase reference signal and an output phase signal of the inverter circuit, and the first driving signal is configured to control the inverter circuit.

Abstract: Disclosed technology of training a third classifier to select between sensitive or non-sensitive categories for a webpage including both sensitive and non-sensitive content. The technology involves collecting webpages as a tentatively labeled dataset A and a dataset B, training a first classifier including at least a first sensitive category classifier and a first non-sensitive category classifier using the tentatively labeled dataset A, referring some dual labelled webpages generated by the first classifier to a curator to curate and resolve label conflict, receiving curated labels from the curator and updating dataset B with the curated labels, training a second classifier using the updated dataset B, referring some dual labelled webpages generated by the second classifier to a curator to curate and resolve label conflict, receiving curated labels from the curator and updating dataset A with the curated labels, and training the third classifier with updated dataset A and updated dataset B.

Abstract: The technology discloses training a classifier to label webpages with categories, extracting from a training database with thousands of webpages tentatively labeled with ground truth categories, dataset A and dataset B; training a classifier using A; and applying it to webpages in B to assign a webpage a label and a classification score. Also disclosed is cleaning B, removing webpages based on evaluation of a decision confidence metric derived from the score assigned for the webpage; and training a second classifier using cleaned B, with second classifier weights initialized independent of trained first classifier weights; and applying the second classifier to webpages in A to assign a webpage the label, score, and decision confidence matrix, and cleaning A, removing webpages from A based on the decision confidence metric. Then combining cleaned A and cleaned B into a combined clean dataset, and training the third classifier using the combined clean dataset.

Abstract: The present disclosure provides a multi-modal encoder processing method and apparatus, a computer device and a storage medium. The method includes: acquiring a pair of mask samples to be processed, the pair of mask samples including a text sample and an audio sample associated with each other, and at least one of the text sample and the audio sample is masked; based on a multi-modal encoder, generating a text encoding feature of the text sample, and generating an audio encoding feature of the audio sample, a linear spectrum feature of the audio sample being fused in the text encoding feature, and a linear word feature of the text sample being fused in the audio encoding feature; and predicting masked mask information according to the text encoding feature and the audio encoding feature, and correcting the multi-modal encoder based on an accuracy of the mask information.

Abstract: Embodiments of the present disclosure provide an audio processing method and apparatus, an electronic device and a storage medium, wherein the method comprises: obtaining first music data and a processing instruction in text form associated with the first music data; extracting, by a music processing model, a first chord progression feature and an audio feature of the first music data, and a text feature of the processing instruction; processing, by the music processing model, the audio feature in accordance with the first chord progression feature and the text feature, to generate second music data; wherein a similarity between a first chord progression feature of the first music data and a second chord progression feature of the second music data is greater than a similarity threshold.

Abstract: In a high-magnification photographing scenario, the electronic device may automatically lock, in response to an operation of the user, a photographing object that requires high-magnification photographing. Then, the electronic device may collect and display a preview image of the photographing object by using a proper zoom ratio, and obtain an initial state and an initial position of the electronic device. In this way, it is unnecessary for the user to manually adjust the zoom ratio, so that jitter caused by the user's manual operation can be reduced.

Abstract: A puncture system, which utilizes intervention navigation technology and a corresponding puncture instrument to operate puncture surgery that features easier operation, higher accuracy and effectiveness, improved safety, and fewer complications, and that has no blind zones during surgery monitoring.

Abstract: A refrigerator includes a box body, a door body, a hinge assembly, and a mounting block. The door body is connected to the box body. The hinge assembly includes a first hinge shaft and a second hinge shaft. The mounting block includes a first trajectory groove and a second trajectory groove. During a process of the door body being rotated to open or close, the first hinge shaft moves relative to the first trajectory groove, and the second hinge shaft moves relative to the second trajectory groove. The first trajectory groove includes a curved groove segment. An end of the curved groove segment extends in a direction proximate to a first side edge, and a distance between the curved groove segment and the door side wall is decreased along a direction from a door rear wall to a door front wall.

Abstract: A coating composition for improving the sag resistance of a gypsum panel or a fibrous panel, the composition including 25 to 100 wt. % alkali metal polyborate silicate inorganic binder, based on the total weight of the dry coating. The inorganic binder includes alkali metal polyborate silicate resulting from combining alkali metal polyborate and alkali metal silicate. Further provided are fibrous panels having a backing side and an opposing facing side including a coating of the inorganic binder, and methods of preparing same. Further provided are gypsum panels having a backing side and an opposing facing side including a coating of the inorganic binder, and methods of preparing same.

Abstract: The present disclosure relates to a puncture system, which can be applied to a puncture operation with electromagnetic combined with ultrasound navigation to achieve a safe, precise, efficient and easy-to-operate puncture procedure.

Abstract: Disclosed is technology for choosing between alternative category labels tentatively assigned to tens of thousands of webpages by a classifier ensemble running on processors, applying the classifier ensemble with a sensitive category classifier, a non-sensitive category classifier, a title and metadata classifier and a heuristic classifier to tens of thousands of webpages. Also disclosed is applying a post processor to outputs of the classifier ensemble and tentatively assigning at least two category labels for non-sensitive categories for webpages; two category labels, automatically determining that at least one but not all of the tentatively assigned category labels is a general label and de-assigning the general label; saving the category label that is not de-selected to the webpage; and distributing the assigned category labels for at least some of the tens of thousands of webpages for use in controlling access to webpages by users on user systems protected using the assigned labels.

Abstract: A refrigerator is provided, which includes a refrigerator body, a pair of door bodies, a turnover beam, a guide seat and a limiting block. The turnover beam is pivotally connected with one of the pair of door bodies, and the turnover beam includes a turnover beam body. The guide seat includes a guide seat body and a guide groove disposed in the guide seat body. The limiting block is rotatably connected with the guide seat body. When at least a portion of the guide block presses the limiting block through the front side opening of the guide groove, the limiting block is configured to pivot and exit the guide groove, and after the guide block enters the guide groove, the limiting block is configured to be reset by pivoting to limit the guide block.

Abstract: A refrigerator includes a box body, a hinge assembly, and a door body. The hinge assembly includes a plate body, a locking hook, and a blocking portion. A free end of the locking hook extends in a direction away from the plate body and is bent in a direction proximate to the plate body, so as to form an opening toward the plate body. In a case where the door body is in a closed state, the blocking portion is located in the opening. During a process from closing to opening of the door body, the locking hook is deformed under force to be disengaged from the blocking portion and unlocked.

Abstract: Various methods, systems, and use cases for securely managing, generating, and controlling access to keys in a service mesh are discussed herein. In various examples, key protection operations include service mesh signing key protection and service mesh communication key protection, for a secure transport session between services such as conducted with mutual transport layer security (mTLS). For instance, such key protection operations may be used to establish communications between the service host and another entity within the service mesh, in a secure transport session, based on use of a private key (secured using a confidential computing technology) in a secure enclave or other secure compute environment to sign one or more keys for the secure transport session.

Abstract: In a high-magnification photographing scenario, the electronic device may automatically lock, in response to an operation of the user, a photographing object that requires high-magnification photographing. Then, the electronic device may collect and display a preview image of the photographing object by using a proper zoom ratio, and obtain an initial state and an initial position of the electronic device. In this way, it is unnecessary for the user to manually adjust the zoom ratio, so that jitter caused by the user's manual operation can be reduced.

Abstract: A refrigerator includes a refrigerator body and a door body. In a case where the door body is opened from a closed state to a first state, a first hinge shaft moves from a first positioning position to a second positioning position along a first positioning trajectory line with respect to a first trajectory slot, and a second hinge shaft moves from a first guiding position to a second guiding position along a first guiding trajectory line with respect to a second trajectory slot. In a case where the door body continues to be opened from the first state to a second state, the first hinge shaft moves from the second positioning position to a third positioning position along a second positioning trajectory line, the second hinge shaft moves from the second guiding position to a third guiding position along a second guiding trajectory line.