Abstract: In a convolutional neural network (CNN) using an encoder-decoder structure for image segmentation, a multi-scale context aggregation module receives an encoded final-stage feature map from the encoder, and sequentially aggregates multi-scale contexts of this feature map from a global scale to a local scale to strengthen semantic relationships of contexts of different scales to improve segmentation accuracy. The multi-scale contexts are obtained by computing atrous convolution on the feature map for different dilation rates. To reduce computation required by the CNN, a channel-wise feature selection (CFS) module is used in the decoder to merge two input feature maps. Each feature map is processed by a global pooling layer followed by a fully connected layer or a 1×1 convolutional layer to select channels of high activation. By subsequent channel-wise multiplication and elementwise summation, only channels with high activation in both feature maps are preserved and enhanced in the merged feature map.

Abstract: The present disclosure discloses a horizontal-axis ocean current power generation device for an underwater vehicle. The power generation device is disposed in a groove of a rotary body of the underwater vehicle, and includes an undercarriage unit, a yawing unit, and a power generation unit. The undercarriage unit can realize elevation and descent of the entire power generation device, and the power generation unit is capable of realizing arbitrary rotation within 360° in a horizontal plane through the yawing unit. The power generation device can actively yaw based on change of an ocean current direction to perform an incident flowing function. The power generation unit respectively drives an outer shaft and an inner shaft to rotate through a front blade and a rear blade that rotate in opposite directions, so as to drive inner and outer rotors of a motor, thereby cutting magnetic induction to generate electric power.

Abstract: The data caching method and a storage medium includes acquiring, by receiving the request of caching service data sent by the UE, the network status data between an SBS corresponding to UE and the UE, the network status data between an MBS corresponding to the SBS and the UE, and the network status data between the UE and a data source of the service data. According to the types of the three pieces of network status data and the service data, three corresponding QoE values are acquired, and whether the cache is configured for the UE is determined according to the three QoE values. The quantification of the three QoE values is realized, and whether the cache is configured for the UE is determined according to the quantized QoE values, thereby achieving the effect of improving the quality of user experience.

Abstract: Systems and methods including one or more processors and one or more non-transitory storage devices storing computing instructions configured to run on the one or more processors and perform receiving a request for a machine learned prediction for an item; determining that a number of entries in item user interaction data of the item is below a predetermined number-of-entries threshold; incorporating category interaction data of a hierarchical category of the item from a predetermined time period into the item user interaction data of the item; determining an initial machine learned prediction for the item using at least the category interaction data, as incorporated into the item user interaction data; and facilitating altering a graphical user interface (GUI) of an electronic device using the initial machine learned prediction for the item. Other embodiments are disclosed herein.

Abstract: In order to solve the problem in the existing conventional water treatment process of low removal efficiency of heavy metal in water, especially lower efficiency for simultaneous removal of heavy metal pollutants during coexisting, a method is provided for removing heavy metal pollutants in water with divalent manganese strengthened ferrate: preparing a ferrate mother liquor having the concentration of 20-10,000 mmol/L; preparing a divalent manganese salt mother liquor having the concentration of 30-10,000 mmol/L; adding the divalent manganese salt mother liquor into water of the heavy metal pollutants; then adding the ferrate mother liquor, and reacting; and then adding a flocculant and precipitating, so that the removal rate of arsenate, chromium, thallium, antimony, chromium and molybdate in water is 90% or more, and the removal rate of heavy metal such as lead and cadmium is 85% or more.

Abstract: A manufacturing method of a template includes: providing a base; forming a photoresist pattern on the base and patterning the base by using the photoresist pattern as a mask, and the forming the photoresist pattern includes: forming a plurality of first patterns spaced apart from each other on the base; forming a first material layer on the plurality of first patterns; patterning the at least one first pattern by using the first material layer as a mask so that the first pattern is formed into at least one first sub-pattern; and removing the first material layer; and the first material layer at least cover one side of at least one of the plurality of first patterns in a direction perpendicular to a surface on which the base is located.

Abstract: Systems and methods including one or more processing modules and one or more non-transitory storage modules storing computing instructions configured to run on the one or more processing modules and perform acts of obtaining a uniform resource locator (URL) of a first webpage that is shown on a graphical user interface and that is external to a website of a retailer (where, in some embodiments, the URL is obtained from a referral website or is entered by a user from a chat window or search box), using a web scraper to extract web text displayed on the first webpage on the graphical user interface, processing the web text displayed on the first webpage on the graphical user interface to determine an interest of a user, using a set of rules to determine items related to the web text displayed on the first webpage on the graphical user interface, and coordinating displaying the items on a second webpage to promote the items as related to the interest of the user, where the second webpage is internal or external

Abstract: Disclosed are methods for treating Set1/COMPASS-associated cancers characterized by expression of Set1B/COMPASS. The methods typically include administering a therapeutic amount of an inhibitor of the Set1B/COMPASS pathway and/or an agonist for a target that is negatively regulated by the Set1B/COMPASS pathway.

Abstract: A method and an apparatus for generating information are provide according to embodiments of the disclosure. A specific embodiment of the method comprises: acquiring to-be-analyzed information according to a target keyword; and inputting the to-be-analyzed information into a pre-established sentiment analysis model to generate sentiment orientation information of the to-be-analyzed information. The sentiment analysis model is obtained through following training: acquiring untagged sample data and tagged sample data; generating tag information corresponding to the untagged sample data using a pre-established tag generation model, and using the untagged sample data and the generated tag information as extended sample data, the tag generation model being used to represent a corresponding relationship between the untagged sample data and the tag information; and obtaining the sentiment analysis model by training using the tagged sample data and the extended sample data.

Abstract: In an embodiment, a method includes programming a control signal that specifies a target resistance and a target voltage in a circuit. The method further includes sending the control signal to at least one transistor configured to control a current flow in the circuit. The method further includes providing, as an output, a signal with the target voltage and target resistance.

Abstract: Provided is a filter membrane modifying method based on metallic oxide particles, relating to the field of membrane method water treatment. The method aims to resolve the problems existing in a common organic membrane or inorganic membrane that an intercepted substance is easily adsorbed on the surface of the membrane and in membrane holes during use so that the membrane holes are clogged, a membrane body is polluted, flux attenuation is quick, the service life of the membrane is shortened, and the backwashing efficiency is low. The method comprises: preparing a solution containing metallic oxide particles, then filtering, by using a filter membrane, the solution containing metallic oxide particles, and forming a metallic oxide particle cover layer on the surface of the filter membrane, i.e., completing the filter membrane modifying method based on metallic oxide particles.

Abstract: Embodiments of a method and a device are disclosed. In an embodiment, a Doherty amplifier module includes a substrate including a mounting surface, and further includes a first amplifier die, a second amplifier die, and a third amplifier die on the mounting surface. The first amplifier die is configured to amplify a first radio frequency (RF) signal along a first signal path, the second amplifier die is configured to amplify a second RF signal along a second signal path, and the third amplifier die is configured to amplify a third RF signal along a third signal path. A side of the first amplifier die including a first output terminal faces a side of the second amplifier die including a second output terminal. The second signal path is parallel to the first signal path, and the third signal path is orthogonal to the first and second signal paths.

Abstract: An amplifier includes a driver stage amplifier transistor and a final stage amplifier transistor, which are integrated in a semiconductor die. The driver stage amplifier transistor has a driver stage input, a driver stage output, and an output impedance, and the driver stage amplifier transistor is configured to operate using a first bias voltage at the driver stage output. The final stage amplifier transistor has a final stage input, a final stage output, and an input impedance. The final stage input is electrically coupled to the driver stage output. The final stage amplifier transistor is configured to operate using a second bias voltage at the final stage output, and the second bias voltage is at least twice as large as the first bias voltage.

Abstract: The invention provides compounds of Formula (I) pharmaceutically acceptable salts, pro-drugs, biologically active metabolites, stereoisomers and isomers thereof wherein the variable are defined herein. The compounds of the invention are useful for treating immunological and oncological conditions.

Abstract: A method of inducing a generalized immune response, includes administering to a subject an immunologically effective amount of one or more isolated immunogenic peptides comprising an amino acid sequence of SEQ. ID NOs: 1-46, thereby inducing a generalized immune response to infection by a pathogen in a subject. A prophylactic or therapeutic composition for inducing a generalized immune response, includes an immunologically effective amount of one or more isolated immunogenic peptides comprising an amino acid sequence of SEQ. ID NOs: 1-46. A method of distinguishing a subject infected with a pathogen from an uninfected subject, includes detecting an antibody in the subject that selectively binds one or more isolated immunogenic peptides comprising an amino acid sequence of SEQ. ID NOs: 1-46, wherein the presence of the antibody indicates that the subject is infected with the pathogen. A general method to discover broadly protective components for a vaccine is also disclosed.

Abstract: An auxiliary sensing method and system based on sensory substitution includes: establishing a discrete horizontal plane included angle and a normal vector included angle in three-dimensional space as well as response functions corresponding to different discrete distances, and establishing an acoustic coding library of an object; obtaining a surrounding environmental video; obtaining object information and category of current environment; establishing a three-dimensional sound field of the object according to the image information, the object information and the three-dimensional coordinate information obtained in combination with the acoustic coding library of the object; and adjusting, for objects of different categories, the three-dimensional sound field of each object according to movement information and variations of images, and playing the three-dimensional sound field of each object in real time.

Abstract: A method including receiving, through a computer network from a search engine, traffic data associated with one or more products, in a time period of a budget period. The method further can include determining an ad-spend amount based at least in part on the traffic data. The method also can include determining a normal range of ad spends for the time period based on a predetermined total allocation amount for the one or more products for the budget period, one or more allocation balancing rules, and/or a spending pattern model. After the normal range is determined, the method further can include detecting an ad-spend anomaly in the time period by monitoring whether the ad-spend amount is outside the normal range. The method additionally can include determining a pacing-control finding that no pacing-control job is being executed. The method further can include transmitting, in real-time through the computer network, an alert to a user in response to the ad-spend anomaly and the pacing-control finding.

Abstract: A system for virtual reality (VR) motion control includes a multi-dimensional motion platform and a thrust universal spherical plain bearing. The VR motion control method includes: setting parameters; calibrating a device; obtaining parameters; receiving a head-mounted display device tracking packet and/or an action control packet; determining whether a real-time angle difference is greater than a set dead zone angle; and calculating a difference between a current angle of a head-mounted display device and a current status of the multi-dimensional motion platform, and feeding the difference into an action control system of the multi-dimensional motion platform, to eliminate the difference. The multi-dimensional motion platform uses single-point bearing or single-point lifting bearing and lever traction-type transmission, thereby realizing functions of swinging and rotating in all directions.

Abstract: Blockchain-based file handling is provided by receiving a data file from a user device, storing the data file to local storage of the blockchain peer, generating a file identifier of the data file, providing the file identifier to the user device, storing the file identifier to a synchronized ledger of the blockchain network, where the synchronized ledger tracks access to the data file, and distributing data of the data file to one or more other blockchain peers of the blockchain network.

Abstract: Systems and methods including one or more processors and one or more non-transitory storage devices storing computing instructions configured to run on the one or more processors and perform acts of storing, for each respective customer of one or more customers, respective customer information in a customer database; predicting, using a machine learning lifetime value (LTV) update model, a respective LTV for each respective customer of the one or more customers; determining an online advertisement for each respective customer of the one or more customers using the respective LTV, as predicted, for each respective customer of the one or more customers; and coordinating displaying the online advertisement for at least a portion of the one or more customers. Other embodiments are disclosed herein.