Abstract: Systems and methods for gradient adversarial training of a neural network are disclosed. In one aspect of gradient adversarial training, an auxiliary neural network can be trained to classify a gradient tensor that is evaluated during backpropagation in a main neural network that provides a desired task output. The main neural network can serve as an adversary to the auxiliary network in addition to a standard task-based training procedure. The auxiliary neural network can pass an adversarial gradient signal back to the main neural network, which can use this signal to regularize the weight tensors in the main neural network. Gradient adversarial training of the neural network can provide improved gradient tensors in the main network. Gradient adversarial techniques can be used to train multitask networks, knowledge distillation networks, and adversarial defense networks.

Abstract: Provided are systems and methods pertaining to analyte detection, including analytes related to SARS-COV-2, such as nucleocapsid protein. The disclosed systems and methods operate by forming a complex between an analyte, a promoter tag, and an anchor and detecting a reaction product that results from the reaction between a reaction substrate and a reaction promoter of the complex. Also provided are systems and methods that allow for quantification of analyte presence by way of monitoring indicator that is displaced by a reaction associate with the analyte.

Abstract: A computer-implemented method, computer program product and computer system to automatically perform file management operations is provided. A processor identifies a plurality of files to monitor. A processor generates tracking attributes for the plurality of files. A processor monitors user interactions with the plurality of files. A processor generates prediction vectors for a plurality of file interactions based on the user interactions with the plurality of files. A processor determines at least one file in the plurality of files with tracking attributes that correlate with at least one prediction vector. A processor performs an operation on the at least one file that corresponds with the at least one prediction vector.

Abstract: An optical imaging lens includes six lens elements. The object-side surface of the first lens element has a convex part in a vicinity of the periphery of the first lens element, the image-side surface of the second lens element has a concave part in a vicinity of the periphery of the first lens element, the object-side surface of the third lens element has a convex part in a vicinity of the optical axis, the image-side surface of the fourth lens element has a convex part in a vicinity of the periphery of the fourth lens element, the image-side surface of the fifth lens element has a concave part in a vicinity of the optical axis, and the sixth lens element has negative refractive power.

Abstract: The present disclosure provides solutions for generating synthetic three-dimensional (3D) objects for testing autonomous vehicle (AV) object detectors, such as AV Light Detection and Ranging (LiDAR) detectors. A process of the disclosed technology can include steps for receiving a first set of object shape data corresponding to a baseline object, permuting the first set of object shape data to generate a second set of object shape data, and providing the second set of object shape data to a differentiable pipeline to generate an output scene, wherein the output scene contains an optimized object that is based on the second set of object shape data. The process can further include steps for determining, using a detector, if the optimized object can be identified in the output scene. Systems and machine-readable media are also provided.

Abstract: Systems and methods are disclosed for automatically predicting a risk score of a user login attempt by receiving a user login attempt and generating a login feature vector associated with the user login attempt. The systems and methods further train a machine learning technique to establish a relationship between the login feature vector and the risk score. The trained machine learning technique is applied to new user login attempts to predict a risk score associated with the login attempt and issue an authentication challenge to the user if the risk score exceeds a predetermined threshold value.

Abstract: A display panel and a display apparatus. The display panel includes a backplane (10) and light-emitting devices (20). The backplane (10) includes: a base substrate (101); pixel driving circuits (102); a first insulating layer (103); conductive portions (104), wherein each of the conductive portions (104) is connected to a corresponding pixel driving circuit (102) via a first connector (CM1) penetrating through the first insulating layer (103); and a second insulating layer (105) located on the side of the first insulating layer (103) and the conductive portions (104) away from the base substrate (101). The light-emitting devices (20) are located on the side of the second insulating layer (105) away from the base substrate (101), wherein an anode (201) of at least one of the light-emitting devices (20) is connected to a corresponding conductive portion (104) via second connectors (CM2) penetrating through the second insulating layer (105).

Abstract: Embodiments of the disclosure generally provide methods of forming a hybrid film stack that may be used as a capacitor layer or a gate insulating layer with a high dielectric constant as well as film qualities for display applications. In one embodiment, a thin film transistor structure include gate, source and drain electrodes formed on a substrate, and an insulating layer formed on a substrate, wherein the insulating layer is a hybrid film stack having a dielectric layer comprising a zirconium containing material disposed on an interface layer formed above or below the gate, source and drain electrodes.

Abstract: A high-speed remote landslide simulation test device with a variable angle includes a support adjustment component, slide plates, and loose leaves. The device first transmits the operation paths of adjusting the slide plate at different angles to the controller, and the personnel operates the control panel to control the support jack, the motor, and the electromagnet to start and stop through the controller, the output shaft of the motor drives the stainless steel threaded rod to rotate, the controller first energizes one electromagnet and disconnects the other three electromagnets, which can only make the stainless steel threaded rod rotate in a fixed position inside the inner threaded pipe, and the three inner threaded pipes follow the rotation direction of the stainless steel threaded rod, thus, the position of the support jack can be moved separately.

Abstract: A method of video decoding includes receiving a coded video bitstream comprising coded information of a current block in a current picture. The coded information is indicative of a bi-directional prediction with a refinement. The refinement determines an offset based on a plurality of calculations, and each calculation includes reference samples of a first reference picture and a second reference picture for the current picture. The method further includes determining that a first calculation in the plurality of calculations uses a first reference sample that is an out of boundary (OOB) prediction sample. An OOB prediction sample of a reference picture is out of a frame boundary of the reference picture. The method also includes performing the refinement that determines the offset based on the plurality of calculations with at least the first reference sample being excluded, and reconstructing the current block according to the bi-directional prediction with the offset.

Abstract: The present disclosure relates to a sgRNA and constructing a dual pig model of severe immunodeficiency and liver injury and use thereof. The method comprises the steps of knocking out RAG2, IL2R? and FAH genes in a porcine fetal fibroblast by using a CRISPR/Cas9 technology, constructing a RAG2?/?/IL2R??/Y/FAH?/? triple-gene edited cloned pig by using a somatic cell nuclear transfer technology, and obtaining a dual pig model of severe immunodeficiency and liver injury through phenotypic analysis and identification. The method overcomes the problems of long production period, low efficiency, irreversible damage, unsatisfactory use in a humanization degree and the like in the existing model construction technology, can realize a batch construction of the dual pig model of severe immunodeficiency and liver injury by a continuous cloning technology, and has great advantages and potential market application prospects in the related fields of tumor biology, cell transplantation, humanized animal models and the like.

Abstract: A video bitstream including a current block in a current picture is received. A set of candidate quantization step sizes of a current block is determined from a plurality of sets of candidate quantization step sizes associated with quantization parameters (QPs). Each of the plurality of sets of candidate quantization step sizes includes a different number of the candidate quantization step sizes. One or more quantization step sizes are determined from the set of candidate quantization step sizes for the current block. Transform coefficients of the current block are determined based on the determined one or more quantization step sizes. The current block is reconstructed based on the determined transform coefficients of the current block.

Abstract: Techniques for establishing communication channels between user devices experiencing network connectivity issues and remote communication systems are described herein. The techniques include the use of a secondary device to act as a proxy, or a “middle man,” to facilitate the communications with the user device. A user device may detect lack of network connectivity, and begin broadcasting advertisement messages that indicate the lack of connectivity. A secondary device may detect the advertisement message, and send a discovery message to a connectivity system indicating that it detected the advertisement message. The connectivity system can provide this information to a remote communication system, and the remote communication system can establish a connection with the secondary device and instruct the secondary device to establish a connection with the user device.

Abstract: A method of decoding includes calculating a first cost value associated with a first sign value and a second cost value associated with a second sign value for a difference weighted bi-prediction, determining a selected sign value from the first sign value and the second sign value according to the first cost value and the second cost value, and determining a first weight value and a second weight value based on a difference between a first reference pixel of a first reference block in a first reference picture and a second reference pixel of a second reference block in a second reference picture for a current pixel in the current block. The method includes reconstructing the current pixel according to a weighted sum of the first reference pixel and the second reference pixel according to the first weight value, the second weigh value and the determined sign value.

Abstract: Systems and methods for distributed machine learning are provided. According to one aspect, a method for distributed machine learning includes obtaining, by an edge device, a static machine learning model from a hub device, computing, by the edge device, an objective function for a dynamic machine learning model based on a relationship between the dynamic machine learning model and the static machine learning model, and updating, by the edge device, the dynamic machine learning model based on the objective function.

Abstract: Aspects of the disclosure provide a method and an apparatus for video encoding/decoding. The apparatus includes processing circuitry for: receiving prediction information of a current coding block in a current picture from a coded video bitstream, the prediction information indicating that the current coding block is coded using a subblock-based temporal motion vector prediction (SbTMVP) mode; deriving multiple displacement vector (DV) candidates by applying multiple DV offset candidates to a fixed DV predictor of the current coding block; comparing a template of the current coding block with each of multiple templates, each template of the multiple templates being located at a position specified by a corresponding one of the multiple DV candidates; calculating a cost value associated with each one of the multiple DV offset candidates based on the comparing; and reordering DV offset indices of the multiple DV offset candidates based on their calculated cost values.

Abstract: The disclosure provides compositions and methods for high-efficiency genome editing in mammals. In some aspects, the disclosure provides guide RNA-retron cassettes, RNA-guided nuclease (e.g., Cas9)-reverse transcriptase fusion protein encoding cassettes, and vectors comprising the cassettes. Also provided are mammalian host cells that have been transfected with the vectors. In other aspects, guide-retron donor DNA molecules are provided. In some other aspects, methods for genome editing in mammals and the screening of genetic loci are provided. In further aspects, methods and compositions are provided for the prevention or treatment of genetic diseases and for other applications. Kits for genome editing and screening are also provided.

Abstract: A data processing method for a log structured merge (LSM) tree includes selecting SST files to be compressed and merged in a current layer and a next layer, sequentially reading the SST files to be compressed and merged in the current layer and the next layer from a first storage device and sequentially writing the SST files in a second storage device, randomly reading the SST files to be compressed and merged from the second storage device into a memory according to key sequence numbers of data blocks included in the SST files to be compressed and merged, and performing compression and merge processing on the SST files to be compressed and merged. Sequential and random read and write speed of the second storage device is higher than that of the first storage device.

Abstract: A vehicle window glass assembly, a vehicle, and a preparation method for the vehicle window glass assembly are provided in the disclosure. The vehicle window glass assembly includes a guide profile, an exterior trim, a fixed glass sheet, and a connection member. The guide profile has a first side face and defines a fixing groove. An opening of the fixing groove is on the first side face. The exterior trim covers an outer surface of the guide profile. The fixed glass sheet includes a fixed portion at an edge of the fixed glass sheet. The fixed portion is received in the fixing groove. The connection member includes a first portion received in the fixing groove. The first portion is connected between a surface of the fixed portion and a groove wall of the fixing groove.

Abstract: Aspects of the disclosure provide methods and apparatuses for video encoding/decoding. In some examples, an apparatus for video decoding includes receiving circuitry and processing circuitry. The processing circuitry receives coded information of a current block in a current picture from a coded video bitstream, the coded information is indicative of applying local illumination compensation (LIC) on the current block according to a reference block in a reference picture. The processing circuitry constructs a parameter set list for the LIC, the parameter set list comprising a plurality of parameter sets for the LIC, and determines a selected parameter set from the parameter set list according to an index that is decoded from the coded video bitstream. The processing circuitry performs the LIC on the current block based on the selected parameter set to generate compensated samples of the current block according to the reference block in the reference picture.