Abstract: A neural network model processing method includes obtaining a first low-bit neural network model through training, where the model includes a first operation layer and a second operation layer. Each operation layer includes at least one operation. Values/a value of a parameter and/or data used for the operation are/is represented by using N bits, and N is a positive integer less than 8. The neural network model processing method further includes compressing the model to obtain a second low-bit neural network model, where the compressed model includes a third operation layer. The third operation layer is equivalent to the first operation layer and the second operation layer, and an operation layer other than the third operation layer in the at least one operation layer is the same as an operation layer other than the first operation layer and the second operation layer in the at least two operation layers.

Abstract: The present disclosure relates to a semiconductor device with isolation and/or protection structures. A semiconductor device can include a substrate, a first transistor and a second transistor, wherein the first transistor and the second transistor are formed on the substrate, and an isolation structure formed on the substrate. The isolation structure can be formed on the substrate between the first transistor and the second transistor. The isolation structure can be configured to isolate the first transistor and the second transistor.

Abstract: Disclosed is a 5-fluorouracil derivative having the molecular structure shown in general formula VI, in which Ra and Rb groups are an alkoxy group or a fluorine-substituted alkoxy group having 1, 2, 3, or 4 carbon atoms, and are mono-, bis-, tri-, tetra- or penta-substituted on a phenyl group; a linking group L1 is an alkyl or alkenyl group having 1, 2, 3, or 4 carbon atoms, a linking group L2 is oxygen, or an alkyl or alkoxy group having 1, 2, 3, or 4 carbon atoms, or an amino acid, or an alkyl group having 1, 2, 3, or 4 carbon atoms containing an amino moiety, or a furyl group, and an X group is O or —NH—. Further disclosed is a method for preparing such a derivative and a use of the same in the treatment of cancer, tumor diseases, and diseases caused by abnormal neovascularization in a human or non-human mammal, and a medicament or a composition containing the 5-fluorouracil derivative.

Abstract: A method for synthesizing a polymer/pigment hybrid latex through sulfur-free and soap-free in-situ reversible addition-fragmentation chain transfer (RAFT) emulsion copolymerization includes the following: dispersing a pigment with a reactive emulsifier to prepare a pigment dispersion; synthesizing an amphiphilic sulfur-free ?-vinyl-terminated polymethyl methacrylate (PMMA) macromolecular RAFT agent in-situ on a surface of pigment particles through catalytic chain transfer polymerization (CCTP); and adding an acrylate monomer dropwise, allowing in-situ sulfur-free RAFT polymerization on a surface of pigment particles through the RAFT agent, and adjusting a structure and a composition of an encapsulated polymer layer to obtain a series of uniformly encapsulated and stably dispersed polymer/organic pigment hybrid latexes.

Abstract: The present disclosure relates to an RF amplifier device including an IC chip including at least one transistor formed on a substrate, at least one operational circuit formed on the substrate and electrically coupled to the transistor, and a port configured to electrically couple the at least one operational circuit with operational circuitry outside the IC chip to adjust operation of the operational circuitry.

Abstract: The present disclosure relates to genetically modified non-human animals that express a human or chimeric (e.g., humanized) CD137, and methods of use thereof.

Abstract: Disclosed are a preparation method of a display panel, a display panel and a displaying device. The display panel comprises a plurality of first-color subpixels, and each first-color subpixel comprises a base, the base comprising a first driving electrode and a second driving electrode; a flat layer disposed on the side, near the first driving electrode and the second driving electrode, of the base; a patterned passivation layer and at least one first electrode disposed on the side, away from the base, of the flat layer, the first electrode being connected with the first driving electrode through via holes penetrating the flat layer; and at least one second electrode disposed on the side, away from the base, of the passivation layer, the second electrode being connected with the second driving electrode through via holes penetrating the passivation layer and the flat layer.

Abstract: A voltage error signal is provided to a PWM controller of a voltage regular and used to produce a PWM signal that drives a power stage of the regulator. When operating in an adapter current limit regulation mode, an adapter current sense voltage, indicative of an adapter current, is compared to an adapter current reference voltage to produce an adapter current error signal. A compensator receives the adapter current error signal and outputs a compensated adapter current error signal. The adapter current sense voltage, or a high pass filtered version thereof, is subtracted from the compensated adapter current error signal to produce the voltage error signal provided to the PWM controller. Alternatively, an input voltage, or a high pass filtered version thereof, is added to the compensated adapter current error signal to produce the voltage error signal.

Abstract: The attention recognition embodied by the method for evaluating the concentration ability of a sensory assessor is organically combined with evaluations for three categories of ranking capability, namely, excellent, good, and poor. Therefore, sensory assessors displaying high sensibility and poor attention form part of the group possessing excellent ranking capability, while sensory assessors exhibiting moderate sensibility and high attention can be found in the group possessing good ranking capability. Furthermore, sensory assessors displaying fair sensibility and high attention can be found in the group with poor ranking capability. This system can identify the concentration ability of assessors, therefore, providing support for the reliability of ranking results.

Abstract: The present disclosure relates to the genetically modified non-human animals that express a human or chimeric (e.g., humanized) T-cell immunoglobulin and mucin-domain containing-3 (TIM-3), and methods of use thereof.

Abstract: A transistor amplifier package includes a base, one or more transistor dies on the base, first and second leads coupled to the one or more transistor dies and defining respective radio frequency (RF) signal paths, and an isolation structure on the base between the respective RF signal paths. The isolation structure includes first and second wire bonds. The first and second wire bonds may have a crossed configuration defining at least one cross point therebetween. Related wire bond-based isolation structures are also discussed.

Abstract: A cam roller type horizontal extrusion cracking system for walnuts, including feeding, cracking and falling devices fixed accordingly on a stand. The feeding device includes a feeding box; an intermittent feeding roller is arranged therein; one side of the roller includes a feeding baffle plate, the other side includes an adjustable feeding scraper blade mechanism; and opposing feeding slots are formed in the roller. The cracking device includes an extrusion box body; movable and fixed tooth-shaped extrusion plates are oppositely mounted therein; one side of the movable plate opposite the fixed includes an extrusion cam; the plates have a plurality of tooth gaps; a walnut passes through the feeding device, falls into a gap between the plates; the cam drives the movable plate to do a periodic reciprocating motion, to synchronously cooperate with the fixed plate to perform extrusion cracking on the walnut.

Abstract: A photolithography alignment method includes: performing alignment measurement of a surface condition of a wafer to obtain alignment information of the wafer; and sectioning the wafer into a plurality of areas to be processed according to the alignment information, and determining photolithography alignment parameters corresponding to each area to be processed.

Abstract: Provided are an array substrate, a display panel, a display apparatus and a preparation method therefor. The array substrate comprises: a base substrate; and multiple pixel units arranged on one side of the base substrate, each of the pixel units comprising: a thin-film transistor and an electroluminescent structure, and a shading structure located between the thin-film transistor and the base substrate, wherein the thin-film transistor comprises: an active layer located on one side, away from the base substrate, of the shading structure; the electroluminescent structure comprises: first electrodes for driving the pixel units; and one of the shading structure and the active layer is a same-layer structure fabricated by the same mask plate as the first electrodes so as to reduce the number of mask procedures required in preparation of an array substrate.

Abstract: The present disclosure relates to genetically modified non-human animals that express a human or chimeric (e.g., humanized) IL15, and methods of use thereof.

Abstract: The present disclosure relates to genetically modified non-human animals that express a human or chimeric (e.g., humanized) IL6R and/or IL6, and methods of use thereof.

Abstract: A dry process connected energy-consuming beam column joint based on a corbel includes a prefabricated concrete corbel column, pre-buried steel plates and connecting steel plates, wherein the corbel section of the prefabricated concrete corbel column is stepped, and a notch section of a prefabricated concrete notch beam is matched with the stepped section of the corbel and is in lap joint to the stepped section; the pre-buried steel plates are separately pre-buried on the upper and lower surfaces of the corbel and the prefabricated concrete notch beam, friction plates are arranged outside the pre-buried steel plates, and slight tooth spaces are arranged on the sides, facing the pre-buried steel plates, of the friction plates; and the connecting steel plates are arranged on the left and right sides of the prefabricated concrete notch beam and the corbel lap joint section.

Abstract: The present disclosure provides an eye-safe laser triangulation measurement system, including a laser transmitter, an optical projection apparatus, a laser beam capturing unit, a cooperation unit, and an image processing apparatus. The laser transmitter is configured to generate point pulse laser of a pulse laser beam. The optical projection apparatus is disposed at a front end of the laser transmitter, and is configured to convert the point pulse laser into a line laser. The laser beam capturing unit is configured to capture reflected light obtained by the line laser at a target position to obtain a laser image. The cooperation unit is configured to control on and off of the laser transmitter, and control capture of the laser beam capturing unit. The image processing apparatus is configured to perform 3D modeling on the laser image to obtain a distance to the target position.

Abstract: The disclosure provides a device for detecting a wavefront error by modal-based optimization phase retrieval using an extended Nijboer-Zernike (ENZ) theory. The detection device includes a point light source (1), a half mirror (2), a lens (3) to be tested, a plane mirror (4) and an image sensor (5). The wavefront error of the component under test is characterized by using a Zernike polynomial, and a Zernike polynomial coefficient is solved based on an ENZ diffraction theory. The present disclosure realizes the one-time full-aperture measurement on the wavefront error of a large-aperture optical component, and can use a partially overexposed image to achieve accurate wavefront error retrieval. Meanwhile, the present disclosure overcomes the contradiction between underexposure and high signal-to-noise ratio (SNR) caused by a limited dynamic range when the image sensor (5) acquires an image. The detection device is simple and does not have high requirements for the experimental environment.

Abstract: A method for preparing isophorone diisocyanate by (1) reacting isophorone with hydrogen cyanide in the presence of a catalyst to obtain isophorone nitrile; (2) reacting the isophorone nitrile obtained in step (1) with ammonia gas and hydrogen in the presence of a catalyst to obtain isophorone diamine; and (3) subjecting the isophorone diamine to a phosgenation reaction to obtain the isophorone diisocyanate, wherein the content of impurities containing a secondary amine group in the isophorone diamine that undergoes the phosgenation reaction in step (3) is ?0.5 wt. The method reduces the content of hydrolyzed chlorine in the isophorone diisocyanate product, improves the yellowing resistance of the product, and the harm due to presence of hydrolyzed chlorine in the product is reduced.