Abstract: Disclosed are a neural network training method, a neural network training device and an electronic device. The neural network training method includes: training a first neural network to be trained by using sample data; determining an indicator parameter of the first neural network in a current training process; determining an update manner corresponding to a preset condition if the indicator parameter meets the preset condition; and updating a parameter of a batch normalization layer in the first neural network based on the update manner. In this way, sparsing of a feature map output by a neural network is implemented, thereby reducing an amount of data to be transmitted and improving computation speed of a chip.

Abstract: A display substrate and a manufacturing method thereof, and a display device. The display substrate includes: a base substrate including a display region and a peripheral region surrounding the display region; a first light emitting element in the display region of the base substrate; and a color film structure on a display side of the first light emitting element. The color film structure includes: a first color film, including only a first pixel color film located in the display region; and a second color film, including a second pixel color film located in the display region and at least partially non-overlapping with the first pixel color film, and a first frame color film located in the peripheral region and surrounding the display region.

Abstract: A genetically modified human beta-glucocerebrosidase (GCase) is disclosed. The genetically modified GCase comprising an amino acid sequence at least 85% identical to SEQ ID NO: 2; and comprising mutations at coordinates L34P, K224N/G, T369E and N370D, where the coordinates correspond to said SEQ ID NO: 2; and capable of catalyzing hydrolysis of a glycolipid glucosylceramide (GlcCer). Pharmaceutical compositions comprising the genetically modified GCase and therapeutic methods of using same are also disclosed.

Abstract: Embodiments pertain to ureteral stents for treatment of partial or full ureteral occlusion by bridging one or more ureteral occlusions, where each stent includes a tubular portion, a first spout portion and a second spout portion. The stent may have one or more tapered edges by having the smallest diameter of the first and/or the second spout portion smaller than the largest diameter of the tubular portion, and/or be of an overall length allowing the ureteral stent to be completely located within a patient's ureter once deployed therein.

Abstract: Disclosed herein are anti-CTLA-4 antibodies possessing superior binding and biological activities relative to ipibmumab and tremelimumab, pharmaceutical compositions comprising such. Also provided herein are therapeutic and diagnostic applications of such anti-CTLA-4 antibodies.

Abstract: A method of reducing a friction coefficient of a surface is disclosed herein, comprising attaching a water-soluble polymer to the surface, and contacting the water-soluble polymer with liposomes, thereby coating the surface with an amphiphilic lipid. Further disclosed herein are solutions comprising a water-soluble polymer attachable to the surface, liposomes, and an aqueous carrier, for reducing a friction coefficient of a surface, and methods utilizing same. Articles of manufacture comprising a substrate coated by a water-soluble polymer which is coated by an amphiphilic lipid are also described, as are uses and methods for treating a synovial joint disorder associated with increased articular friction.

Abstract: The present application discloses a circuit for generating spread-spectrum synchronous clock signal. The circuit includes a frequency detector comprising a fraction controller configured to compare an input signal of a first frequency with a feedback signal of a second frequency in a loop of feedback to generate a first control signal and a second control signal alternately for determining a control word to track the first frequency and a phase-shift controller configured to register n levels for the first control signal and the second control signal to introduce n phase delays for changing a fraction part of the control word randomly to provide a broadened boundary. The circuit also includes a digitally controlled oscillator configured to generate a synthesized periodic signal based on a base time unit, the first frequency, and the control word, with the second frequency being locked within the broadened boundary of the first frequency.

Abstract: Multi-agent reinforcement learning may be used for rider order-dispatching via matching the distribution of orders and vehicles. Information may be obtained. The information may include a plurality of vehicle locations of a plurality of vehicles, a plurality of ride orders, and a current time. The obtained information may be input into a trained model. The trained model may be based on Kullback-Leibler divergence optimization and independent agents under a guidance of a joint policy. A plurality of order-dispatching tasks may be generated for the plurality of vehicles to fulfill.

Abstract: A display panel and a display device are provided, the display panel includes a gate driving circuit, a first pixel module, and a second pixel module, the first pixel module includes a first pixel circuit, and the second pixel module includes a second pixel circuit; the gate driving circuit respectively provides corresponding gate driving signals for the first pixel circuit and the second pixel circuit; in the first display region, the number of columns of the pixel circuit included by the first pixel module is greater than the maximum number of columns of the pixel circuit included by the second pixel module in the second display region; the display panel further includes a loading structure arranged in a non-display region; the load structure is electrically connected to the gate wiring; the gate wiring is wiring between the gate driving circuit and the second pixel circuit for providing the corresponding gate driving signals to the second pixel circuit.

Abstract: The present disclosure provides an energy-efficient process for preparing nitrogen and oxygen for a glass melting furnace. A device required by the process includes a filter, a turbine air compressor, an air pre-cooling unit, alternately used molecular sieve adsorbers, an electric heater, a main heat exchanger, a rectifying tower I, a main condenser-evaporator I, a rectifying tower II, a main condenser-evaporator II, a rectifying tower III, a main condenser-evaporator III, a supercooler, an expander I and an expander II. The three rectifying towers are used to prepare a low-pressure nitrogen product and an oxygen product with a certain pressure at the same time. The oxygen product with a certain pressure is used for oxygen-enriched combustion for the glass melting furnace, and the low-pressure nitrogen product is used as shielding gas of a tin bath.

Abstract: A nanopaper and a fabricating method thereof, a method of graphic processing and an electronic device are provided. The nanopaper includes: a transparent substrate, wherein the transparent substrate includes a micro-nano-fiber; a plurality of pressure sensing units, wherein the pressure sensing units are located on one side of the transparent substrate, and resistances of the pressure sensing units are able to vary with deformation of the transparent substrate; and a plurality of leads, wherein the plurality of leads are connected to the pressure sensing units, and are configured to be able to separately output a sensing signal of each of the plurality of pressure sensing units.

Abstract: A method includes: obtaining a plurality of first signals corresponding to a vehicle and a plurality of second signals corresponding to a plurality of candidate carpool combinations each comprising one or more unassigned transportation orders, wherein: the plurality of first signals comprise a current time, a location of the vehicle at the current time, and one or more static features corresponding to the vehicle, the plurality of second signals comprise timestamps, origins, and destinations of the unassigned transportation orders, and the vehicle has an on-going transportation order at the current time; inputting the plurality of first and second signals to a trained machine learning model; and obtaining, from an output of the trained machine learning model, a utility score of each of the plurality of candidate carpool combinations.

Abstract: The present disclosure relates to a mirror display device, an image display method, an electronic device and a storage medium are provided and related to the technical field of optics. The mirror display device comprises: an optical switch array (101) comprising multiple optical switches, the optical switches being used for obtaining, in a first state, a first light beam based on incident light; a reflection layer (102) arranged on one side of the optical switch array (101) and used for receiving and reflecting the first light beam incident to the reflection layer (102); and a control module (103) for determining a reflection region and a non-reflection region, controlling the optical switch corresponding to the reflection region to be in the first state, and controlling the optical switch corresponding to the non-reflection region to be in a second state that is different from the first state.

Abstract: A solar cell can include a silicon semiconductor substrate; an oxide layer on a first surface of the silicon semiconductor substrate; a polysilicon layer on the oxide layer; a diffusion region at a second surface of the silicon semiconductor substrate; a dielectric film on the polysilicon layer; a first electrode connected to the polysilicon layer through the dielectric film; a passivation film on the diffusion region; and a second electrode connected to the diffusion region through the passivation film.

Abstract: A speed change device for shifting the gear of a transmission. The transmission includes a first shifter shaft, a first shifter element associated to the first shifter shaft and arranged for movement in an axial direction of the first shifter shaft for actuating a first set of gears. It further comprises a second shifter shaft and a second shifter element associated to the second shifter shaft and arranged for movement in an axial direction of the second shifter shaft for actuating a second set of gears. One of said first and second shifter shafts is hollow and the other of said first and second shifter shafts is partly arranged coaxially inside the hollow shaft.

Abstract: Method of predicting a response of a subject to food is disclosed. The method comprises: selecting a food to which a response of the subject is unknown; accessing a first database having data describing the subject but not a response of the subject to the selected food; accessing a second database having data pertaining to responses of other subjects to foods, the responses of the other subjects including responses of at least one other subject to the selected food or a food similar to said selected food; and analyzing the databases based on the selected food to estimate the response of the subject to the selected food.