Abstract: The disclosed computer-implemented method may include receiving, by a computing device, an original video to relight. Additionally, the method may include predicting, by the computing device using a de-lighting model trained on a hybrid dataset of lighting-rich data and motion-rich data, an albedo video corresponding to the original video. The method may also include generating, by the computing device using a relighting model trained on the hybrid dataset, a relit video based on the albedo video under a specified lighting condition based on an input high dynamic range (HDR) map. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: Methods for motion-controllable video diffusion include extracting optical flow fields from an input video and computing warped noise by iteratively warping noise between consecutive frames using the optical flow fields. The iteratively warping includes (i) re-Gaussianizing expanded pixel regions by sampling fresh Gaussian noise, and (ii) aggregating contracted pixel regions by merging noise particles and renormalizing variance to preserve spatial Gaussianity. An output video is generated by initializing a diffusion process with the warped noise and iteratively denoising to produce temporally coherent output frames. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A color film substrate includes a first base substrate, a color film layer disposed at one side of the first base substrate, a protective layer disposed at one side of the color film layer away from the first base substrate, and a plurality of first spacers disposed corresponding to the plurality of first grooves. In the non-display region, a plurality of first grooves are provided at one surface of the protective layer away from the first base substrate. A part of the first spacer is disposed in the first groove.

Abstract: The present disclosure provides a cleaning robot, a cleaning module, a cleaning assembly, a base and a cleaning system, wherein the cleaning robot includes: a robot apparatus, configured to carry the cleaning module; and a first replacement mechanism, configured to enable the robot apparatus to be loaded with the cleaning module, and replace a cleaning module carried on the robot apparatus with the loaded cleaning module. When the cleaning robot is loaded with a new cleaning module, the loaded new cleaning module replaces the old cleaning module originally carried on the cleaning robot, in this way, the assembling and disassembling of the cleaning modules on the cleaning robot can be performed synchronously.

Abstract: A neural network-based error compensation method for mass transfer of mini-light-emitting diode (Mini-LED) chips includes the following steps. (S1) An automated optical re-inspection result is obtained as a first result. (S2) The first result is sorted and normalized to obtain a second result. (S3) Nearest-neighbor interpolation is performed on a Mini-LED chip with a transfer status identifier being abnormal, and a differential of path variables of each Mini-LED chip is calculated. (S4) A multi-layer neural network model is defined. A loss function is constructed. A weight of the multi-layer neural network model is updated with the loss function, until no overfitting is observed. (S5) A chip transfer path is generated and input into the multi-layer neural network model to obtain a predicted transfer error value, and mass transfer of the Mini-LED chip is performed based on the predicted transfer error value.

Abstract: The technology of this application relates to a neural network model training method, an image processing method, and an apparatus in the artificial intelligence. The training method includes each of at least one first accelerator training a neural network model based on at least one training sample. Before forward computation at an ith layer is performed, different parameters of the ith layer are obtained locally and from another accelerator, to obtain a complete model parameter of the ith layer. According to the method in this application, storage pressure of the first accelerator can be reduced.

Abstract: An error-correctable planar stretching expansion device for chips includes a first film expansion mechanism, a second film expansion mechanism, a film pressing frame, a scanning camera, and an error correction mechanism. The error correction mechanism includes a correction disk, ejector pins, and a drive mechanism. The first film expansion mechanism and the second film expansion mechanism are employed to stretch a blue film globally. The scanning camera detects positions of chips on the blue film. The correction disk locally secures regions of the blue film where the chips are non-deviated by means of suction through holes. The ejector pins lift regions of the blue film where the chips are deviated and unsecured, thereby performing localized stretching of the blue film, thereby correcting the positions of the chips on the blue film and enabling precise adjustment of chip alignment.

Abstract: A chip pitch control method is provided, in which a first position matrix is generated, and a target matrix of the chips is constructed according to the first position matrix; a first displacement vector set of the chips is constructed; a second displacement vector set is generated, and a unit displacement vector each chip corresponding to a pinhole is obtained by quadratic spline interpolation in combination with the second displacement vector set; an objective function is iterated, and pinhole coordinates are simultaneously adjusted; the number of pinholes is adjusted, and compared with the number of chips; the target matrix is adjusted according to comparison results, and step (S2) or step (S6) is performed; a needle is controlled to pierce a back film according to the adjusted pinhole coordinates. A chip pitch control device is also provided.

Abstract: A press liquid pump includes an outer pipe and an inner pipe disposed in the outer pipe. A vacuum pump head is disposed on the outer pipe. A liquid outlet pipe of the vacuum pump head is disposed outside the outer pipe. A pressing portion of the vacuum pump head is disposed in the outer pipe. A first opening is defined in the outer pipe. A first end of the inner pipe is closed, and a second end of the inner pipe defines a second opening. When the inner pipe is disposed in the outer pipe, the inner pipe is inserted into the outer pipe, and the inner pipe is connected with the pressing portion of the vacuum pump head. When the inner pipe is pressed towards the vacuum pump head, liquid in the inner pipe is pumped to an outside of the outer pipe.

Abstract: Disclosed is a method for resource recycling of high-salinity wastewater treated by a solar-enhanced electrochemical process. The method includes: step S1: subjecting the high-salinity wastewater to an ion membrane electrolysis process to generate a strongly oxidizing salt solution A; step S2: mixing the strongly oxidizing salt solution A with organic wastewater to obtain a mixed solution, and activating and degrading the mixed solution with a photothermal coupled activation degradation system to obtain an activated and degraded solution; step S3: degrading the activated and degraded solution through an electrochemical oxidation degradation component to obtain a high-salinity solution; step S4: evaporating and concentrating the high-salinity solution to obtain a high-salinity concentrate; step S5: subjecting the high-salinity concentrate to the ion membrane electrolysis process to generate a strongly oxidizing salt solution B; and Step S6: repeating step S2 to step S5.

Abstract: A color film substrate includes a first base substrate, a color film layer disposed at one side of the first base substrate, a protective layer disposed at one side of the color film layer away from the first base substrate, and a plurality of first spacers disposed corresponding to the plurality of first grooves. In the non-display region, a plurality of first grooves are provided at one surface of the protective layer away from the first base substrate. A part of the first spacer is disposed in the first groove.

Abstract: Techniques pertaining to enhancements to ranging and reporting in wireless communications are described. An apparatus (e.g., an initiator) performs a one-to-many contention-based ranging operation by: (a) generating a one-to-many initiator report; and (b) transmitting the one-to-many initiator report to a responder. The one-to-many initiator report contains a resolvable private address (RPA) hash calculated using an identity resolving key (IRK) of one of the responder.

Abstract: The disclosed computer-implemented method may include receiving, by a computing device, multi-view flat-lit performance data of a subject. Additionally, the method may include rendering, by the computing device, a dynamic sequence of novel-view flat-lit images of the subject based on a deformable three-dimensional Gaussian splatting (3DGS) model. The method may also include providing the rendered dynamic sequence of flat-lit images as input to a diffusion-based relighting model trained on the multi-view flat-lit performance data of the subject. Furthermore, the method may include generating, by the computing device using the diffusion-based relighting model, a relit sequence of the subject under a specified lighting condition. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: An anode current collector and a preparation method therefor, a battery cell, a battery, and an electric device. The anode current collector includes a metal substrate and metal nanoparticles located on at least part of a surface of the metal substrate; and the metal substrate includes a first metal element, and the first metal element is made of a material the same as the metal nanoparticles.

Abstract: A synchronization method includes: acquiring radio frame data from a sending node; performing a search of a synchronization signal in the radio frame data in the radio frame to determine location information of the synchronization signal in the radio frame data; determining a time calibration parameter of the receiving node according to the location information of the synchronization signal and a frame structure of the radio frame; and calibrating a local clock of the receiving node according to the time calibration parameter and the local clock of the receiving node, so that the receiving node is synchronized with the sending node.

Abstract: A method for processing a semiconductor structure and a method for forming a word line structure are provided. The method for processing the semiconductor structure includes: providing a semiconductor structure including a groove and a metal layer located in the groove, where an edge position of a top surface of the metal layer is higher than a center position of the top surface of the metal layer; enabling the semiconductor structure to be in a rotating state; and performing at least one metal surface planarization process on the semiconductor structure, so that the top surface of the metal layer after being processed is more planar than the top surface of the metal layer before being processed. Each of the at least one metal surface planarization process includes: etching the top surface of the metal layer by a first reagent; and cleaning the semiconductor structure by a second reagent.

Abstract: The present disclosure provides an LED module manufacturing method and an LED module. The LED module manufacturing method includes: selecting a lamp bead substrate and connecting a plurality of groups of LED chips to the lamp bead substrate; forming an encapsulation adhesive layer on the lamp bead substrate by an encapsulation adhesive to encapsulate the plurality of groups of LED chips; preparing a metal thin film layer on the encapsulation adhesive layer and preparing a metal wire grid, which has grid wires extending in the same direction, by the metal thin film layer; cutting the lamp bead substrate based on each group of LED chips; repeatedly performing the above steps; arranging the at least two types of LED lamp beads, which have metal wire grids with grid wires extending in different directions, in an array on a module substrate to obtain an LED module.

Abstract: An electrolyte, a lithium metal battery, a preparation method thereof, and an electric apparatus are provided. The electrolyte includes a lithium salt, an organic solvent, and an electrolyte additive. The additive has a structure of R-M-X, where R is a hydrocarbon group, M is a metal having a standard electrode potential higher than that of lithium, and X is an anion or a hydrocarbon group. M can form an intermetallic compound or alloy solid solution with lithium. The lithium metal battery includes a negative electrode having a current collector coated with a lithiophilic layer. The lithiophilic layer contains M, which also has a standard electrode potential higher than that of lithium and can form an intermetallic compound or alloy solid solution with lithium. This configuration enhances the lithiophilicity of the negative electrode surface, promotes uniform lithium nucleation, and improves the cycling performance of the lithium metal battery.

Abstract: A system for managing data availability for database transactions is disclosed. The system is programmed to receive source data and extract a plurality of events from the source data. The system is programmed to create a halt record corresponding to each event of the plurality of events. The system is programmed to resolve conflicts among one or more halt records. The system is programmed to receive a request for a specific database transaction related to a specific database item from a client device. The system is programmed to determine whether any halt record is in effect for the specific database item. The system is programmed to perform the specific database transaction in response to determining no halt record is in effect. The system is programmed to transmit a result of the specific database transaction in response to the request to the client device.