Abstract: A federated learning method using synonym includes: sending a general model to every client device by a moderator; performing a training procedure by each client device, wherein the training procedure includes: encoding private data into a digest; training a client model according to the private data, the digest and the general model, and sending the digest and a client parameter of the client model to the moderator; determining an absent client device of the plurality of client devices by the moderator; generating a synonym of the digest corresponding to the absent client device by a synonym generator; training a replacement model according to the synonym and the digest corresponding to the absent client device by the moderator; and performing an aggregation to update the general model by the moderator according to the replacement model and the client parameter of every client devices except the absent client device.

Abstract: The present disclosure provides automatic cleaning system and method. The automatic cleaning system includes a control unit, a Micro Control Unit (MCU), a cleaning device and a sensor device. The control unit is configured to determine a cleaning type for cleaning a protective cover of a camera based on an image captured by the camera and the measured temperature and humidity on the outer surface of the protective cover. The control unit is also configured to generate a cleaning instruction based on the cleaning type. The control unit is also configured to send the instruction to the MCU. The MCU is configured to control the cleaning device to clean the protective cover of the camera in accordance with the cleaning instruction in response to receiving the instruction sent from the control unit.

Abstract: Systems and methods for assessing similarity of documents are provided. Embodiments of the systems and methods include extracting a reference document text from a reference document, extracting an archived document text from an archived document, and quantifying the reference document and the archived document. The systems and methods may also include determining a document similarity value of the quantified reference document and the archived document. Determining the document similarity value includes calculating a set of vector similarity values for a set of combinations of a reference document text vector and an archived document text vector, and calculating the document similarity value, including a sum of the plurality of vector similarity values.

Abstract: Provided is a cleaning apparatus, including a liquid supply tank, a liquid storage tank, a first cleaning component, a first rib, disposed on the first cleaning component, for being in contact with a member to be cleaned; where the first rib has two opposite walls protruding above a top surface of the first cleaning component, and the two opposite walls extend along a length direction of the first cleaning component; and a liquid supply port, supplying cleaning liquid in the liquid supply tank to the first cleaning component, where the liquid supply port faces down toward a liquid input end of the first cleaning component; and where liquid injected from the liquid input end flows toward the member to be cleaned and cleans the member to be cleaned under a guidance of the two opposite walls, and is collected to the liquid storage tank.

Abstract: An underwater cleaning robot system includes an underwater robot and an overwater control platform, the underwater robot includes a cleaning robot and a driving robot, and the cleaning robot is carried on the driving robot; and the cleaning robot includes a motor and a pump, the pump is connected to the motor and driven by the motor to generate high-pressure fluid, and the overwater control platform communicates with the cleaning robot and the driving robot to control the cleaning robot and the driving robot. The overwater control platform and the underwater robot are connected by a thin zero-buoyancy cable, and the working range is over 1000 meters.

Abstract: A method for document similarity analysis. The method includes generating a reference document content identifier for a reference document, including identifying frequently occurring terms in reference document content, encoding each frequently occurring term in a term identifier and combining the term identifiers to form the reference document content identifier associated with the reference document. The method also includes obtaining at least one document similarity value by comparing the reference document content identifier to a set of archived document content identifiers stored in a document repository.

Abstract: An image capture device is disclosed that includes a body defining a peripheral cavity and a door assembly that is movable between an open position and a closed position to close and seal the peripheral cavity. The door assembly includes a door body; a slider that is supported by the door body for axial movement between a first position and a second position; a biasing member that is configured for engagement (contact) with the slider; a door lock including a stop that is configured for engagement (contact) with the biasing member; and a sealing member that is fixedly connected to the door lock.

Abstract: Provided are a battery winding method, a battery winding system, a battery and an electrical device. The battery winding method includes: winding; photographing a picture of a current winding layer; acquiring position data of a first point and of a second point according to the picture of the current winding layer, converting the position data of the first point to obtain the converted position data of a converted first point using a preset conversion matrix corresponding to the current winding layer based on the number of the current winding layers and calculating data of displacement between the first and second electrode plates based on the converted position data of the converted first point and the position data of the second point, and determining if the data of displacement is within a threshold value scope, if so, returning to wind a next winding layer, and if not, sending an alarm.

Abstract: Embodiments of the present disclosure relate to a data storage method, a data read method, an electronic device, and a computer program product. The data storage method includes: acquiring identifiers of a predetermined number of trusted nodes, the predetermined number being associated with a secret sharing scheme to be used; generating a predetermined number of shared portion sets from input data by using the identifiers respectively as parameters of the secret sharing scheme; and respectively storing the predetermined number of shared portion sets in the predetermined number of trusted nodes according to a correspondence relationship between the predetermined number of shared portion sets and the predetermined number of trusted nodes. With the technical solution of the present disclosure, input data can be efficiently protected with trusted nodes by using a secret sharing scheme, and the confidentiality and integrity of the input data can be guaranteed.

Abstract: An invention provides a chimeric antigen receptor. The chimeric antigen receptor includes an extracellular domain including a heavy chain variable region, a light chain variable region of a single chain antibody fragment and CD8 hinge region; a transmembrane domain including an immune co-stimulator transmembrane domain; and an intracellular domain including an immune co-stimulator intracellular segment and CD3? chain. According to the embodiments of the invention, the chimeric antigen receptor can specifically recognize the tumor cells expressing the specific antigen and achieve the specific killing effect against the tumor cells expressing the high specific antigen.

Abstract: The present invention relates to a CPR posture monitoring model and system, the model being configured to: based on human skeleton point data extracted from CPR moves, at least compute arm-posture angle data and GC matching angle data related to the CPR moves, and determine whether the CPR moves are qualified by comparing arm-posture angle data and GC matching angle data to a CCP qualification threshold; wherein human skeleton point data are extracted from first move data collected by a first optical component and second move data collected by a second optical component simultaneously, wherein an included angle between collection directions of the two optical components ranges between 30 and 90 degrees. By constructing and applying a CPR posture monitoring model defined by data like GC matching angle, evaluation result has higher accuracy.

Abstract: In some examples, an electronic device comprises a wireless transceiver to transmit a packet and a processor coupled to the wireless transceiver. The processor is to update a running average of a radiation level over a time period. The running average is based on a quantity of packets transmitted by the wireless transceiver during the time period and a radiation level associated with the quantity of packets. The processor is to determine whether the running average exceeds a threshold and to control transmission power for the packet based on the determination and based on a destination of the packet.

Abstract: An operating system upgrade method and device, and a storage medium. The method includes: obtaining an operating system upgrade package that includes a second partition table and operating system upgrade data, where address configurations of partitions with a same name are consistent in the second partition table and a first partition table; triggering a first restart of the electronic device; updating a partition table of the electronic device to the second partition table in the recovery mode; triggering a second restart of the electronic device, where after the second restart, the electronic device loads data of a basic partition, a first static partition, and a dynamic partition to run the first operating system; and upgrading an operating system of the electronic device according to the operating system upgrade data, and upgrading the first operating system to a second operating system.

Abstract: A drowsy driving detection method comprises: acquiring a side face image of a currently seated driver collected by a camera module; performing face recognition on the side face image to obtain side face feature parameters, and determining, according to the side face feature parameters, whether an ID file corresponding to the currently seated driver exists in a driver ID library; and if yes, periodically acquiring a side face image of the driver in the current period collected by the camera module, obtaining eye movement feature parameters of the driver in the current period according to the side face image of the current period, and determining whether the driver is driving while drowsy according to a comparison result between the eye movement feature parameters of the current period and the normal eye movement feature parameters of the driver.

Abstract: The present invention discloses a preparation process of multi-component spherical alloy powder, which adopts a plasma rotation electrode process (PREP) method to prepare the multi-component spherical alloy powder. The multi-component alloy includes at least one of refractory metals and compounds thereof, specifically including tungsten, molybdenum, tantalum, niobium, rhenium, tungsten carbide, tantalum carbide and the like.

Abstract: A non-diffusion type photodiode is described and has: a substrate, a buffer layer, a light absorption layer, an intermediate layer, and a multiplication/window layer. The buffer layer is disposed on the substrate. The light absorption layer is disposed on the buffer layer. The intermediate layer is disposed on the light absorption layer and has a first boundary, wherein the intermediate layer is an I-type semiconductor layer or a graded refractive index layer. The multiplication/window layer is disposed on the intermediate layer and has a second boundary, wherein in a top view, the first boundary surrounds the second boundary, and a distance between the first boundary and the second boundary is greater than or equal to 1 micrometer. The non-diffusion type photodiode can reduce generation of dark current.

Abstract: A button for use with an image capture device in an underwater environment that includes a movable plunger configured to cause actuation of the image capture device and upper and lower components collectively defining an internal cavity that is configured to receive the plunger. The upper and lower components are configured and connected such that actuation of the image capture device is prevented until a threshold pressure is applied to the upper component that is greater than external water pressure applied in the underwater environment. The upper component includes at least one opening that is configured to allow water to enter the internal cavity to modify an internal pressure within the internal cavity so as to reduce the threshold pressure required to actuate the image capture device in the underwater environment.

Abstract: An anode passivation slurry of anti-dendritic lithium and a method of preparation are provided. The method comprises the steps of dissolving a divalent copper metal compound and a non-ionic polymer to obtain a first solution, dissolving trimesic acid to obtain a second solution, and mixing the first and second solutions to obtain a copper-based metal-organic framework. The dried precursor are mixed with ionic liquid, which has contained a first lithium salt, and then dried to obtain an anion impregnated copper-based metal-organic framework. Thereafter, an anion impregnated copper-based metal-organic framework, a second lithium salt, polymer materials, and a second solvent are mixed to obtain the anode passivation slurry. The anode passivation slurry homogenizes the concentration of conduction of lithium ions and improves ionic conductivity, reducing the formation of lithium dendrites, and improving the cycle life of batteries. A battery with the anode passivation slurry dried on an anode is also disclosed.