Abstract: A numerical characterization method for the dispersion state of carbon nanotubes based on fractal dimension is invented. In this method, a SEM image of the dispersion state of carbon nanotube is obtained first, and then is binarized by the ImageJ software to extract the boundary of individual carbon nanotubes or carbon nanotubes agglomerates, and thereby calculating the fractal dimension of the processed image with the assistance of the box-count algorithm. The value of fractal dimension represents quantitatively the abundant information contained in the dispersion state of carbon nanotubes, which is capable of realizing the numerical characterization of the dispersion state of carbon nanotubes. The invention quantifies the dispersion state of carbon nanotubes, and provides a powerful strategy for controlling, comparison and prediction of macro-properties of carbon nanotube based composites.

Abstract: A dual-output port charging circuit includes a primary switching circuit, a secondary first switching circuit, a secondary second switching circuit, a transformer connected to the three circuits, and a controller. The controller receives and samples an output voltage and an output current of the secondary second switching circuit, the output voltage and the output current are compared with a reference value, and a compensation is made, so as to eventually control an on-off time of a switch in the primary switching circuit; and the controller receives and samples an output current of the primary switching circuit, a zero-cross delay is calculated through a zero detection, so as to control an on-off time of a rectification switch in the secondary second switching circuit. One controller is used to flexibly control voltages of two direct current output ports with accurate voltage regulation and strong antijamming capability.

Abstract: Methods and systems of annotation densification for semantic segmentation are disclosed herein. In one example embodiment, such a method includes obtaining image information, obtaining coarse annotation information, performing an image matting operation based upon the image information and based at least indirectly upon the coarse annotation information, and applying an already-trained Convolutional Neural Network (ConvNet) semantic segmentation model in relation to the image information.

Abstract: Presented are systems and methods for 3D reconstruction from videos using an unsupervised learning framework for depth and normal estimation via edge-aware depth-normal consistency. In embodiments, this is accomplished by using a surface normal representation. Depths may be reconstructed in a single image by watching unlabeled videos. Depth-normal regularization constrains estimated depths to be compatible with predicted normals, thereby, yielding geometry-consistency and improving evaluation performance and training speed. In embodiments, a consistency term is solved by constructing depth-to-normal layer and normal-to-depth layers within a deep convolutional network (DCN). In embodiments, the depth-to-normal layer uses estimated depths to compute normal directions based on neighboring pixels. Given the estimated normals, the normal-to-depth layer may then output a regularized depth map. Both layers may be computed with awareness of edges within the image.

Abstract: A computer system receives, from a client device, a virtual bill request including user identification information of a user associated with the client device and then activates a virtual bill and associates the virtual bill with the client device, the virtual bill having a predefined effective time period. After receiving data corresponding to one or more items added to the virtual bill within the predefined effective time period through respective close proximity data communication between the client device and one or more respective terminals, the computer system determines a value based on the data corresponding to the one or more items and sends a verification request to the client device. Upon receipt of verification information responsive to the verification request, the computer system initiates a transfer of the value from an account associated with the user in accordance with a determination that the verification information is valid.

Abstract: The present disclosure describes a font retrieval system that utilizes a multi-learning framework to develop and improve tag-based font recognition using deep learning neural networks. In particular, the font retrieval system jointly utilizes a combined recognition/retrieval model to generate font affinity scores corresponding to a list of font tags. Further, based on the font affinity scores, the font retrieval system identifies one or more fonts to recommend in response to the list of font tags such that the one or more provided fonts fairly reflect each of the font tags. Indeed, the font retrieval system utilizes a trained font retrieval neural network to efficiently and accurately identify and retrieve fonts in response to a text font tag query.

Abstract: Provided herein are compounds, preferably thyroid hormone receptor beta (THR beta) agonist compounds, compositions thereof, and methods of their preparation, and methods of agonizing THR beta and methods for treating disorders mediated by THR beta.

Abstract: An instrument for dynamically observing the evolution behavior of the slope topography. The dual hosts emit parallel, equal-height and overlapped laser planes from different directions, and project the laser planes onto a slope landform. The image acquisition device shoots the slope landform with the projected laser lines at an angle perpendicular to the laser-planes. Then a video screenshot can be formed and imported into a computer to form a three-dimensional map. Finally, the volume, slope gradient and other parameters of the slope body can be obtained. As the knobs on the mechanically micro-adjustable host of the topography meter are turned, the laser-planes can become precisely equidistant and parallel. A novel topography meter is disclosed, which can conveniently calibrate the spacing and angle of the linear lasers and accurately observe all landforms including the local deep trenches on an eroding slope.

Abstract: The present disclosure provides a collimator. The collimator includes a slice module, a filter module, and a support and protection module. The slice module and the filter module are detachably connected with the support and protection module. The rays strike on the collimator and pass through the slice module and the filter module in order. The slice module is configured to adjust the fan-beam width of the rays striking on the collimator. The filter module is configured to make the rays pass through different filters to implement different types of scanning.

Abstract: The present application provides a shift register circuit, a gate driving circuit including the shift register circuit, and a driving method applied to the shift register circuit. The shift register circuit includes an input sub-circuit, an output sub-circuit, an output reset sub-circuit, and a first capacitor, wherein the first capacitor is connected between the pull-up node and the second clock signal terminal, and configured to maintain a high level at the pull-up node through the second clock signal input at the second clock signal terminal. The shift register circuit further includes a second capacitor connected between the pull-down node and a first voltage input terminal, and configured to pull down a level at the pull-down node through a reverse bias voltage input at the first voltage input terminal during a blanking time after a frame of scanning ends.

Abstract: A method for presenting a data handling method on a graphical user interface includes identifying a type of service associated with a pending data handling process, sending one or more user characterizations associated with a user's identity to a server for performing the data handling process, receiving data handling capabilities corresponding to one or more data handling methods for performing the identified type of service, determining one or more candidate data handling methods to be displayed on a graphical user interface (GUI) of the client computing device, initiating display of the candidate data handling methods on the client computing device GUI based on the data handling capabilities, performing the pending data handling process with a particular data handling method selected from the candidate data handling methods, and sending a performance result associated with the pending data handling process.

Abstract: A network device generates a successive joint channel estimation based on interference cancellation processes to eliminate or reduce colliding interferences (colliders) at a receiver end. A selection component selects a strongest interference collider from among pseudo noise sequences operating as colliders. A 2 by 2 matrix successive joint channel estimation component performs a successive 2 by 2 matrix joint channel estimation with an initial correlation metric between the strongest interference collider and each remaining interference collider in a first iteration of a set of iterations. A correlation metric component generates an updated correlation metric vector based on the 2 by 2 matrix joint channel estimation. Another 2 by 2 matrix joint channel estimation is then performed with the updated correlation metric vector between a next strongest interference collider and remaining interference colliders of the plurality of interference colliders at a second iteration of the set of iterations.

Abstract: Disclosed are various embodiments for automated determination of application privileges. An application testing service executes an application in a testing computing environment that replicates a production computing environment. The application is given a set of permissions in the testing computing environment. A set of use cases for the application is performed while the application is executing in the testing computing environment. A set of operations invoked by the application to perform the set of use cases is determined. A subset of the set of permissions is generated that corresponds to a minimal set of permissions necessary to invoke the set of operations.

Abstract: The present invention provides an aqueous composition for coating substrates comprising: one or more emulsion copolymer, the emulsion copolymer being a copolymerization product of a monomer mixture comprising a monomer of formula I, wherein R1 is hydrogen or methyl, X is a divalent organic alkylene oxide group, n is an integer ranging from 1 to 50 and R2 is hydrogen or an aliphatic or aromatic hydrocarbon group having from 1 to 20 carbon atoms, and a phosphate surfactant. When incorporated into a coating composition, the coating composition show both improved stain-resistance properties and anti-clogging properties.

Abstract: An objective of the present invention is to provide a method for determining and controlling a robot to move straightly on a sloping plane to solve a technical issue that a deviation of a conventional robot from a predetermined straight path is difficult to detect when the robot moves on a sloping plane and cannot be corrected, and moving on the straight path is difficult to guarantee.

Abstract: Aspects of the disclosure provide methods and apparatuses for video decoding. In some examples, an apparatus includes receiving circuitry and processing circuitry. The processing circuitry decodes prediction information of a current block in a current picture from a coded video bitstream. The prediction information is indicative of an affine model in a merge mode. Further, the processing circuitry obtains, from a buffer, motion information of bottom locations in a neighboring block that is adjacent of the current block in the current picture. Then the processing circuitry determines, based on the motion information of the bottom locations in the neighboring block, parameters of the affine model that is used to transform between the block and a reference block in a reference picture that has been reconstructed. Further, the processing circuitry reconstructs at least a sample of the current block based on the affine model.

Abstract: An electronic device at a base station end includes a processing circuit, the processing circuit being configured to: configure, in response to request signalling from a user equipment, an aperiodic beam-forming reference signal relevant to a first beam group for the user equipment, wherein the first beam group is determined by a base station according to channel state information periodically fed back by the user equipment; generate downlink control information, so as to indicate that the user equipment feeds back beam selection information according to the aperiodic beam-forming reference signal; determine, according to the beam selection information, one or a plurality of candidate beams and one or a plurality of corresponding second pre-coding codebooks; and determine an effective pre-coding codebook based on the one or multiple second pre-coding codebooks.