Abstract: A system for training parameters of a neural network includes a processing node with a processor reconfigurable at a first level of configuration granularity and a controller reconfigurable at a finer level of configuration granularity. The processor is configured to execute a first dataflow segment of the neural network with training data to generate a predicted output value using a set of neural network parameters, calculate a first intermediate result for a parameter based on the predicted output value, a target output value, and a parameter gradient, and provide the first intermediate result to the controller. The controller is configured to receive a second intermediate result over a network, and execute a second dataflow segment, dependent upon the first intermediate result and the second intermediate result, to generate a third intermediate result indicative of an update of the parameter.

Abstract: The invention discloses a parametric system for testing a capability of auditory spatial positioning and a method of providing a parametric minimum audible angle. The system includes: an host computer system calculating spatial movement data and communicating with a slave computer; a spatial movement system realizing control of any spatial position through circumferential movement, radial movement and vertical movement; a real-time three-dimensional display system built into the host computer system; an audio system randomly generates a type of audio data of a sound source and a sound source with equal-difference decibels; and a test system for matching human spatial positioning results to sound source spatial data and providing an objective evaluation. The invention further provides a method of providing a parametric minimum audible angle for a capability of auditory spatial positioning, the invention improves the control accuracy.

Abstract: A stain resistant coating composition includes: urethane acrylate core-shell particles including (1) a polymeric acrylic core; and (2) a polymeric shell including a urethane linkage. The polymeric shell is formed from a reaction mixture including an isocyanate and a polyol. The polyol is substantially free of a six or more consecutive methylene group chain and/or the polyol is formed from monomers comprising a monomer content including at least 10 wt % substituted polyol and/or substituted polyacid and/or at least 25 wt % polyol or polyacid containing cyclic content, based on the total weight of the monomers forming the polyol.

Abstract: The present disclosure belongs to the field of preparation technology and provides an ultralimit alloy and a preparation method therefor. The ultralimit alloy comprises an alloy matrix. A bonding layer and a ceramic layer are successively deposited on a surface of the alloy matrix. The alloy matrix includes one of a magnesium alloy matrix, an aluminium alloy matrix, a titanium alloy matrix, an iron alloy matrix, a nickel alloy matrix, a copper alloy matrix, a zirconium alloy, and a tin alloy. For an ultralimit magnesium alloy, an ultralimit aluminium alloy, an ultralimit nickel alloy, an ultralimit titanium alloy, an ultralimit iron alloy and an ultralimit copper alloy, the bonding layer is a composite bonding layer, the ceramic layer is a composite ceramic layer, and the outside of the composite ceramic layer is further successively deposited with a reflecting layer, a catadioptric layer, an insulating layer and a carbon foam layer.

Abstract: A liquid crystal cell includes a first substrate, a second substrate and a liquid crystal layer disposed between the first substrate and the second substrate. The first substrate includes a first base, and an electrochromic layer disposed on a side of the first base. The second substrate includes a second base, and a plurality of spacers arranged at intervals and disposed on a side of the second base proximate to the first substrate. An orthographic projection of at least one spacer of the plurality of spacers arranged at intervals on the first base is within a range of an orthographic projection of the electrochromic layer on the first base.

Abstract: A crosslinking composition includes a compound having at least two functional groups that are each independently represented by Chemical Structure (I): X is an oxygen, sulfur, or nitrogen; R1 is an alkyl group, an aryl group, or an alkylaryl group; R2, R3, and R4 are each independently an alkyl group, an aryl group, an alkylaryl group, or a hydrogen; R5 is an alkyl group, an aryl group, an alkylaryl group, or a hydrogen; z is 0 when X is oxygen or sulfur and z is 1 when X is nitrogen; and when a double bond is formed between a carbon atom bonded to R3 and an adjacent nitrogen, m is 0, and when a single bond is formed between the carbon atom bonded to R3 and the adjacent nitrogen, m is 1.

Abstract: The present disclosure provides a light-adjusting glass and a smart vehicle window. The light-adjusting glass has a transmittance adjustment region and an encapsulation region; the light-adjusting glass includes: a first and second substrate opposite to each other, and a dye liquid crystal layer between the first and second substrates in the transmittance adjustment region, and a frame sealant in the encapsulation region; the first substrate includes a first base and a first electrode layer on a side of the first base proximal to the dye liquid crystal layer; the second substrate includes a second base and a second electrode layer on a side of the second base proximal to the dye liquid crystal layer; a conductive structure is provided in the frame sealant; a first and second voltage transmission structures electrically insulated from each other are on the first base.

Abstract: A calibrating method is provided including the following steps. A type of a first sensor and a type of a first sensor carrier are determined according to an external shape of a first object. The first sensor is carried by the first sensor carrier, and a relative coordinate of the first object is measured by the first sensor. The relative coordinate of the first object is compared with a predetermined coordinate of the first object to obtain a first object coordinate error, and the first object coordinate error is corrected. After the first object coordinate error is corrected, the first object is driven to perform an operation on a second object or the second object is driven to perform the operation on the first object. A calibrating system is also provided.

Abstract: The present disclosure provides a smart glass and a light adjusting method, a light adjusting device and a non-transitory computer storage medium. The smart glass includes a first glass module, a touch control layer and a control unit. The first glass module includes a first substrate and a second substrate opposite to each other, and a first light adjusting layer between the first and second substrates. The light transmittance of the first light adjusting layer is adjusted by an electric field between the first substrate and the second substrate. The touch control layer is provided in at least one of the first substrate and the second substrate. The control unit is electrically coupled to the touch control layer to receive a touch control signal output by the touch control layer and adjust the electric field between the first substrate and the second substrate according to the touch control signal.

Abstract: A dimming panel and a manufacturing method thereof are provided. The dimming panel includes: a first base substrate and a second base substrate opposite to the first substrate; a first electrode on the first base substrate; a second electrode on the second base substrate; and a liquid crystal layer between the first and second base substrate. The first electrode includes a plurality of first electrode strips arranged at intervals in a first direction and a plurality of second electrode strips arranged at intervals in the first direction, the plurality of first electrode strips are located in a first electrode layer, the plurality of second electrode strips are located in a second electrode layer on a side of the first electrode layer away from the first base substrate. An orthographic projection of a combination of first and second electrode electrode strips on the first base substrate is an integrated plane without gaps.

Abstract: A crosslinking composition includes a compound having at least two functional groups that are each independently represented by Chemical Structure (I): X is an oxygen, sulfur, or nitrogen; R1 is an alkyl group, an aryl group, or an alkylaryl group; R2, R3, and R4 are each independently an alkyl group, an aryl group, an alkylaryl group, or a hydrogen; R5 is an alkyl group, an aryl group, an alkylaryl group, or a hydrogen; z is 0 when X is oxygen or sulfur and z is 1 when X is nitrogen; and when a double bond is formed between a carbon atom bonded to R3 and an adjacent nitrogen, m is 0, and when a single bond is formed between the carbon atom bonded to R3 and the adjacent nitrogen, m is 1.

Abstract: Roughly described, a system for data parallel training of a neural network on multiple reconfigurable units configured by a host with dataflow pipelines to perform different steps in the training CGRA units are configured to evaluate first and second sequential sections of neural network layers based on a respective subset of training data, and to back-propagate the error through the sections to calculate parameter gradients for the respective subset. Gradient synchronization and reduction are performed by one or more units having finer grain reconfigurability, such as an FPGA. The FPGA performs synchronization and reduction of the gradients for the second section while the CGRA units perform back-propagation through the first sequential section. Intermediate results are transmitted using a P2P message passing protocol layer. Execution of dataflow segments in the different units is triggered by receipt of data, rather than by a command from any host system.

Abstract: A data processing system comprises a pool of reconfigurable data flow resources and a runtime processor. The pool of reconfigurable data flow resources includes arrays of physical configurable units and memory. The runtime processor includes logic to receive a plurality of configuration files for user applications. The configuration files include configurations of virtual data flow resources required to execute the user applications. The runtime processor also includes logic to allocate physical configurable units and memory in the pool of reconfigurable data flow resources to the virtual data flow resources and load the configuration files to the allocated physical configurable units. The runtime processor further includes logic to execute the user applications using the allocated physical configurable units and memory.

Abstract: The present disclosure provides an optical assembly, a liquid crystal display device, and an electronic equipment. The optical assembly includes: a linear polarizer, a half-wave plate, and a quarter-wave plate stacked in sequence. An absorption axis of the linear polarizer is substantially perpendicular to a first direction, and the first direction is parallel to a surface of the linear polarizer; an angle between an in-plane slow axis of the half-wave plate and the first direction is in a range of 100° to 110°; an angle between an in-plane slow axis of the quarter-wave plate and the first direction is in a range of 160° to 170°.

Abstract: Provided are an Ethereum Virtual Machine-based transaction processing method and apparatus, a device and a medium, which relate to the field of computer technologies and, in particular, to blockchain technologies and are applicable to the field of cloud computing. An implementation scheme includes: acquiring, by a virtual machine instance running in a blockchain node, a to-be-processed transaction request; in a process of executing the to-be-processed transaction request, generating, by the virtual machine instance, a data access request for target access data and transmitting the data access request to an interface module, where the data access request includes a data read request and/or a data write request; and calling, by the interface module, a blockchain access interface for a data storage space of a blockchain according to the data access request to access the target access data and feeding back an access result to the virtual machine instance.

Abstract: Provided are an Ethereum Virtual Machine-based transaction processing method, a device, a program and a medium. An implementation scheme includes: acquiring, by a virtual machine instance running in a blockchain node, a to-be-processed transaction request; in a process of executing the to-be-processed transaction request, generating, by the virtual machine instance, a data access request for an intra-contract variable and transmitting the data access request to an interface module; performing, by the interface module, instruction conversion according to the data access request to determine an intra-contract variable access interface corresponding to a function of the data access request; and calling, by the interface module, the intra-contract variable access interface for a contract space in a data storage space of a blockchain to access the intra-contract variables in the contract space and feeding back an access result to the virtual machine instance.

Abstract: Provided are an Ethereum Virtual Machine-based transaction processing method and apparatus, a device, a program and a medium. An implementation scheme includes: acquiring, by a virtual machine instance running in a blockchain node, a to-be-processed transaction request; in a process of executing the to-be-processed transaction request, generating, by the virtual machine instance, a data access request for target access data and transmitting the data access request to an interface module; performing, by the interface module, instruction conversion according to the data access request to determine a blockchain access interface corresponding to a function of the data access request; and calling, by the interface module, the blockchain access interface for a data storage space of a blockchain to access the target access data and feeding back an access result to the virtual machine instance, where the data storage space includes an account space and a contract space.

Abstract: Provided are an Ethereum Virtual Machine-based transaction processing method and apparatus, a device, a program and a medium, which relate to the field of computer technologies and, in particular, to blockchain technologies. An implementation scheme includes: acquiring, by a virtual machine instance running in a blockchain node, a to-be-processed transaction request; generating, by the virtual machine instance, a data access request for target access data and transmitting the data access request to an interface module in a process of executing the to-be-processed transaction request; performing, by the interface module, instruction conversion according to the data access request to determine a blockchain access interface corresponding to a function of the data access request; and calling, by the interface module, the blockchain access interface for a data storage space of a blockchain to access the target access data and feeding back an access result to the virtual machine instance.

Abstract: Provided are a display substrate and a control method therefor, a display panel, and a display device. The display substrate is disposed in the display panel and includes a first substrate and multiple mode switching electrodes and multiple data lines disposed on the first substrate; wherein, the mode switching electrode is disposed on a side of the data line away from the first substrate and insulated from the data line, and an orthographic projection of the mode switching electrode on the first substrate covers part of an orthographic projection of the data line on the first substrate; and the mode switching electrode is configured to switch display modes of the display panel, wherein the display modes include a normal mode and a privacy mode.

Abstract: A measuring program compiling device and a measuring program compiling method thereof are provided. The method includes analyzing a first measuring program to obtain a plurality of first measuring parameters corresponding to the first measuring program, and converting the first measuring parameters into a plurality of second measuring parameters corresponding to a plurality of planning operations; generating a plurality of standardized measuring parameters according to the second measuring parameters and a plurality of computer aided design (CAD) image parameters; receiving a plurality of parameter input operations corresponding to the planning operations to update the second measuring parameters; generating a standardized measuring program corresponding to a CAD file according to the standardized measuring parameters; and converting the standardized measuring program into a target measuring program executed on a target measuring device according to specification data of the target measuring device.