Abstract: In certain aspects, a three-dimensional (3D) memory device includes a first semiconductor structure, a second semiconductor structure, and a bonding interface between the first semiconductor structure and the second semiconductor structure. The first semiconductor structure includes an array of memory cells, a first peripheral circuit of the array of memory cells, and a polysilicon layer between the array of memory cells and the first peripheral circuit. The first peripheral circuit includes a first transistor. The second semiconductor structure includes a second peripheral circuit of the array of memory cells. The second peripheral circuit includes a second transistor.

Abstract: The present invention provides a composition comprising an antiviral agent and a C—X—C chemokine receptor type 4 (CXCR4) antagonist, and methods for using the same. Compositions and methods of the invention can be used to treat viral infection, immune disorders (e.g., rheumatoid arthritis), pulmonary fibrosis, or any other clinical condition associated with abnormal expression (e.g., overexpression and/or upregulation) of CXCR4. Compounds of the invention can also be used in stem cell therapeutics.

Abstract: A cross-coupling control method for moving beam of gantry machine tool is disclosed, which relates to the technical field of CNC machine tool control. The cross-coupling control method includes: Step 1: establishing a crossbeam dynamics model considering a ram on the crossbeam, and at the same time simplifying the model for an observer design; step 2: realizing a PID control parameter adjustment of motors at both ends in accordance with a method of parameter tuning of a unilateral servo control system; using a servo system with the same control parameters to jointly drive the crossbeam up and down, realizing the synchronous control and realizing PID control parameter tuning for both end motors. The disclosure not only reduces the synchronization error caused by the torsion of the crossbeam, but also solves the synchronization error caused by the asymmetric load on both sides, and improves the system robustness and stability.

Abstract: Systems, methods, and processor-readable media for elastic allocation of resources for deep learning jobs are described. A machine-learning-as-a-service (MLaaS) of a cloud computing system includes an elastic training module which includes resource allocator for allocating resources to training jobs that optimizes overall estimated time to completion (ETC) for all training jobs received by the system and uses node-based resource allocation. The elastic training module may realize a combination of high resource utilization, short training times, and low queueing delay relative to existing approaches, thereby potentially enabling the realization of higher profits for a cloud computing system which provides MLaaS to users (i.e. customers). An improved user interface is described, enabling users to specify a range of resources to elastically allocate to the user's training job, and/or informing users of training time saved through the use of elastic resource allocation.

Abstract: An information processing method and apparatus, a terminal, and a storage medium. The information processing method comprises: determining first content in response to a first operation event of a first control in a first document (S11); and adding the first content to the first document on the basis of content information and type information of the first content (S12). The type information comprises first type information and/or second type information, the second type information having an association with the first type information. In the described method, first content can be added to a first document according to content information and type information of the first content, so as to distinguish different ways of adding the first content.

Abstract: Provided are systems for analyzing a relational database using embedding learning that may include at least one processor programmed or configured to generate one or more entity-relation matrices from a relational database and perform, for each entity-relation matrix of the one or more entity-relation matrices, an embedding learning process on an embedding associated with an entity. When performing the embedding learning process on the embedding associated with the entity, the at least one processor is programmed or configured to generate an updated embedding associated with the entity. Computer-implemented methods and computer program products are also provided.

Abstract: Various exemplary aspects are directed to mineral wool plant substrates comprising mineral wool fibers bound by a binder formed from an aqueous binder composition comprising a cross-linking agent comprising at least two carboxylic acid groups; a polyol component having at least two hydroxyl groups; a nitrogen-based protective agent; and from 0.05 wt. % to 0.7 wt. % of a non-ionic surfactant. Methods of manufacturing such fiber-based plant substrates can include collecting a plurality of inorganic fibers on a substrate; applying an aqueous binder composition to the collection of inorganic fibers, forming binder-coated inorganic fibers; and curing the aqueous binder composition at a temperature less than 510° F., thereby forming a fiber-based plant substrate, wherein said aqueous binder composition is free of formaldehyde.

Abstract: In certain aspects, a three-dimensional (3D) memory device includes a first semiconductor structure, a second semiconductor structure, a third semiconductor structure, and a first bonding interface between the first semiconductor structure and the second semiconductor structure. The first semiconductor structure includes an array of memory cells. The second semiconductor structure includes a first peripheral circuit of the array of memory cells. The first peripheral circuit includes a first transistor. A third semiconductor structure includes a second peripheral circuit of the array of memory cells. The second peripheral circuit includes a second transistor. The first semiconductor structure, the second semiconductor structure, and the third semiconductor structure are stacked over one another.

Abstract: An assembled subway station and a construction method thereof are provided. The assembled subway station includes a plurality of components combined in a ring shape, comprising a bottom plate component, two bottom corner components, two side wall components, a middle plate component, two top corner components, and a top plate component. Connection surfaces of each component are each provided with a concave-convex structure being matched with another concave-convex structure on a relevant connection surface of an adjacent component. Butt joint positions of two adjacent components are filled with sealing structures. The embedded grooves of two adjacent components are oppositely arranged, and connecting members for exerting opposite fastening forces on the two adjacent components are installed in the oppositely arranged embedded grooves.

Abstract: An audio visual haptic signal reconstruction method includes first utilizing a large-scale audio-visual database stored in a central cloud to learn knowledge, and transferring same to an edge node; then combining, by means of the edge node, a received audio-visual signal with knowledge in the central cloud, and fully mining semantic correlation and consistency between modals; and finally fusing the semantic features of the obtained audio and video signals and inputting the semantic features to a haptic generation network, thereby realizing the reconstruction of the haptic signal. The method effectively solves the problems that the number of audio and video signals of a multi-modal dataset is insufficient, and semantic tags cannot be added to all the audio-visual signals in a training dataset by means of manual annotation. Also, the semantic association between heterogeneous data of different modals are better mined, and the heterogeneity gap between modals are eliminated.

Abstract: An image processing method includes: obtaining DCE magnetic resonance images corresponding to a plurality of time points for a same detection target; determining average pixel grayscale values of images of a same lesion region in the DCE magnetic resonance images of the plurality of time points respectively; determining a time to peak according to the average pixel grayscale values corresponding to the plurality of time points; and generating a first-stage time-intensity image before the time to peak and a second-stage time-intensity image after the time to peak respectively according to the DCE magnetic resonance images and the time to peak. The first-stage time-intensity image and the second-stage time-intensity image are 3D images. A pixel grayscale value of each pixel in the first-stage time-intensity image and the second-stage time-intensity image represents a change rate of blood supply intensity and reflects a severity level of a lesion corresponding to the lesion region.

Abstract: A smart acoustic information recognition-based welded weld impact quality determination method and system, comprising: controlling a tip of an ultrasonic impact gun (1) to perform impact treatment on a welded weld with different treatment pressures, treatment speeds, treatment angles and impact frequencies, obtaining acoustic signals during the impact treatment, calculating feature values of the acoustic signals, and constructing an acoustic signal sample set including various stress conditions; marking the acoustic signal sample set according to impact treatment quality assessment results for the welded weld; establishing a multi-weight neural network model, and using the marked acoustic signal sample set to train the multi-weight neural network model; obtaining feature values of welded weld impact treatment acoustic signals to be determined, inputting the feature values into the trained multi-weight neural network model, and outputting determination results for welded weld impact treatment quality to be det

Abstract: Embodiments provide a read operation circuit, a semiconductor memory, and a read operation method. The read operation circuit includes: a data determination module configured to read read data from a memory bank, and determine whether to invert the read data according to the number of bits of low data in the read data to output global bus data for transmission through a global bus and inversion flag data for transmission through an inversion flag signal line; a data receiving module configured to determine whether to invert the global bus data according to the inversion flag data to output cache data; a parallel-to-serial conversion circuit configured to perform parallel-to-serial conversion on the cache data to generate output data of the DQ port; and a precharge module configured to set an initial state of the global bus to High.

Abstract: A method for constructing a large-span station by two-wing open type semi-covered excavation and semi-reverse construction, is divided into three stages of excavation. First, excavate a first-stage inner small foundation pit, then excavate a second-stage annular foundation pit within the first-stage retaining piles and outside the range of the first-stage inner small foundation pit, and finally excavate a third-stage semi-covered excavation foundation pit below the first-stage inner small foundation pit and the second-stage annular foundation pit. By setting graded retaining piles, middle upright post piles, middle top plates, local waist beams and local concrete supports in the soil-rock combination strata, so that the force transfer between the foundation pit enclosure and the station main body structures and the underlying rock layer is clear and reliable, and a stable frame structure is achieved.

Abstract: The present invention relates to application of gossypol and its optical isomers to preparation of an anti-Coronavirus drug. Specifically disclosed is application of pharmaceutically acceptable salts, solvates, isotopes, stereoisomers, stereoisomer mixtures, and tautomers of the gossypol and its optical isomers to preparation of a drug for preventing and/or treating a disease caused by a Coronavirus. The Coronavirus is MERS-CoV, SARS-CoV, or SARS-CoV-2. It is found in the present invention for the first time that the gossypol and its optical isomers can inhibit the activity of Coronavirus 3CL proteases to achieve an anti-Coronavirus effect. The half maximal inhibitory concentrations of gossypol and (?)-gossypol to SARS-CoV proteases and SARS-CoV-2 3CL proteases are all lower than 10 ?M. The anti-Coronavirus 3CL protease effect is good. Therefore, the gossypol and its optical isomers have the potential for use in the preparation of a drug for preventing and/or treating novel Coronavirus infections.

Abstract: Described are a system, method, and computer program product for dynamic node classification in temporal-based machine learning classification models. The method includes receiving graph data of a discrete time dynamic graph including graph snapshots, and node classifications associated with all nodes in the discrete time dynamic graph. The method includes converting the discrete time dynamic graph to a time-augmented spatio-temporal graph and generating an adjacency matrix based on a temporal walk of the time-augmented spatio-temporal graph. The method includes generating an adaptive information transition matrix based on the adjacency matrix and determining feature vectors based on the nodes and the node attribute matrix of each graph snapshot.

Abstract: Provided is a system for analyzing a multivariate time series that includes at least one processor programmed or configured to receive a time series of historical data points, determine a historical time period, determine a contemporary time period, determine a first time series of data points associated with a historical transaction metric from the historical time period, determine a second time series of data points associated with a historical target transaction metric from the historical time period, determine a third time series of data points associated with a contemporary transaction metric from the contemporary time period, and generate a machine learning model, wherein the machine learning model is configured to provide an output that comprises a predicted time series of data points associated with a contemporary target transaction metric. Methods and computer program products are also provided.

Abstract: The present invention relates to an aqueduct leakage repairing method, a formwork is provided within an aqueduct body; a reinforcing layer is formed by using the formwork through casting at inner sides of the aqueduct body; the formwork comprises side formwork plates, bottom formwork plates, and axillary angle formwork plates; the side formwork plates are arranged at inner vertical walls of the aqueduct body, the bottom formwork plates are arranged on a bottom of the aqueduct body, the bottom formwork plates are connected with the side formwork plates through the axillary angle formwork plates; and the axillary angle formwork plates are each provided with a plurality of ribs. The aqueduct leakage repairing method effectively reinforces penetration damage of the aqueduct, and the construction process is simple since the formwork is free from dismantling after construction.

Abstract: In one example, a circuit comprises: a controller, an electrostatic discharge (ESD) circuit, and a driver circuit. The controller has a driver control output. The ESD circuit has a driver control input and an ESD output, the driver control input coupled to the driver control output. The driver circuit has a driver input and a driver output, the driver input coupled to the ESD output.