Abstract: The present disclosure provides a method and apparatus for a route navigation, which relate to the technical fields of computer vision, image processing, augmented reality and the like. An implementation solution is: performing an augmented reality navigation on a current route based on real-time acquired GPS information of a terminal; acquiring SLAM (simultaneous localization and mapping) coordinates of the terminal; converting, in response to determining that the GPS information meets a preset positioning difference condition, the SLAM coordinates into calculated geographic coordinates; and performing an augmented reality navigation on the current route based on the calculated geographic coordinates.

Abstract: A semiconductor package comprises a lead frame, a chip, and a molding encapsulation. The lead frame comprises a top plate, a plurality of drain pads, a plurality of slanted sections, a gate pad, and a plurality of source pads. The top plate of the lead frame comprises a thicker region and a thinner region. Each slanted section of the plurality of slanted sections connects a respective drain pad of the plurality of drain pads to the top plate. A respective side surface of each drain pad of the plurality of drain pads is exposed from a side surface of the molding encapsulation. A respective bottom surface of each drain pad of the plurality of drain pads is exposed from a bottom surface of the molding encapsulation. A top surface of the thicker region is exposed from a top surface of the molding encapsulation.

Abstract: A white light emitting device with an efficiency of at least 230 lm/W at a blue LED chip input current density from 10 to 60 mA/mm2, preferably in the range from 15 to 40 mA/mm2 and more preferably in the range from 20 to 30 mA/mm2. The device comprises a substrate, at least one string of blue LED chips mounted on the substrate and a phosphor material composition. Said phosphor material composition comprises a narrow band red phosphor which generates light with a peak emission wavelength in a range from 625 nm to 635 nm. The weight percentages of the narrow band red phosphor are between 33 to 49 wt. % for a CCT of from 4000 to 6500K or in an amount of from 60 to 70 wt. % for a CCT of from 2700 to 3500K CCT.

Abstract: A traction battery assembly includes, among other things, a plurality of battery cells disposed along an axis, an endplate at an axial end of the plurality of battery cells, and a corrugated area of the endplate. The corrugated area has a plurality of corrugations that flatten to accommodate expansion of the plurality of battery cells along the axis.

Abstract: An unreal engine graphic construction method includes the following steps: receiving data to be displayed; determining a number N of basic graphics contained in a graphic to be constructed according to the data to be displayed, wherein, N>1; calling an underlying function at a blueprint layer to draw N of the basic graphics in sequence, wherein the underlying function is used for creating the basic graphics.

Abstract: Disclosed are a method and apparatus for compressing a topology recognition model, an electronic device, and a medium. The method includes: pruning a to-be-compressed model to obtain a pruned model, and training the pruned model to obtain a trained pruned model; quantifying a bit quantity of a weight parameter of each neural network layer in the trained pruned model from a first bit quantity to a second bit quantity to obtain a highly-quantified pruned model, and training the highly-quantified pruned model to obtain a trained highly-quantified pruned model; using the trained highly-quantified pruned model as the to-be-compressed model, continuously pruning and quantifying the to-be-compressed model until a compressed topology recognition model is obtained, and deploying the compressed topology recognition model on an Internet of Things in Power Systems (IOTIPS) terminal, to reduce a size of the topology recognition model while ensuring the topology recognition model meets an accuracy requirement.

Abstract: The present disclosure provides an image processing method and device. The image processing method includes: performing, by an encoder and a first model, multiple iterations on an initial image to obtain a target image feature corresponding to the initial image; and performing, by a second model, image reconstruction based on the target image feature to obtain a reconstructed image of the initial image, both of the first model and the second model being neural networks for image reconstruction, wherein in the multiple iterations, an image feature extracted by the first model in the image reconstruction and an output image of the first model are feedback information for the encoder to assist the encoder in encoding the initial image.

Abstract: An image generation system is described. The system comprises a neural network model configured to perform a diffusion process to generate a set of multi-view images from a same input prompt. The set of multi-view images have a same subject from different view orientation. The neural network model comprises a self-attention layer configured to relate pixels across the set of multi-view images.

Abstract: The present disclosure provides an expression driving method and apparatus, and a training method and apparatus of an expression driving model. The expression driving method includes acquiring a first video; and inputting the first video into a pre-trained expression driving model to obtain a second video. The expression driving model is trained based on a target sample image and a plurality of first sample images. A facial image in the second video is generated based on the target sample image. A gesture expression feature of the facial image in the second video is the same as a gesture expression feature of a facial image in the first video.

Abstract: A method for controlling an interference torque for a vehicle is provided. The method is applied to a new energy vehicle including an electric motor and an engine, and includes steps of: arbitrating between a pedal torque of a driver and an interference torque required by ESP; performing an initial allocation on the electric motor and/or the engine in response to the pedal torque when the vehicle is in a hybrid drive mode, to meet an engine torque request while ensuring that the engine is operated at an optimal operation point; and determining, based on the initial allocation, whether the motor is capable of fully responding to the arbitrated torque, if so, controlling the motor to respond to the arbitrated torque in priority, otherwise controlling the engine and the motor to cooperatively respond to the arbitrated torque.

Abstract: The present disclosure provides a method of generating an image with metallic texture, and a method of training a metallic texture image generation model. The method of generating an image with metallic texture includes: acquiring a first video; and inputting the first video into a pre-trained metallic texture image generation model to obtain a second video. Each frame of images in the second video is an image with metallic texture. The metallic texture image generation model is trained based on a plurality of first sample images and second sample images with metallic texture corresponding to each first sample image.

Abstract: The present disclosure provides a video generation method and device. The video generation method includes: extracting a first image feature from a first image; obtaining a plurality of intermediate image features by means of nonlinear interpolation according to the first image feature and a second image feature, wherein the second image feature is an image feature of a second image; and performing image reconstruction by means of an image generation model based on the first image feature, the second image feature, and the plurality of intermediate image features, so as to generate a target video, wherein the target video is used for presenting a process of a gradual change from the first image to the second image.

Abstract: A method includes accessing a plurality of weight matrices of a machine learning model. The method also includes, for each weight matrix, decomposing the weight matrix into a U matrix, an S matrix, and a V matrix using singular value decomposition. The S matrix is a diagonal matrix, and a singular group corresponds to each element in the S matrix. The method further includes, for each weight matrix, determining an importance score of each singular group. The importance score of the singular group represents a change in loss if the singular group is removed from the machine learning model. The method also includes, for each weight matrix, ranking the singular groups across the plurality of weight matrices based on the importance scores. In addition, the method includes, for each weight matrix, identifying one or more of the singular groups to prune based on the ranking of the singular groups.

Abstract: The present invention relates to a thin-film composite (TFC) membrane composition comprising macrocycles. The invention also relates in part to a method of fabricating a TFC membrane and to a method of using the TFC membrane to separate a desired liquid or gas from a liquid or gas mixture.

Abstract: This application relates to a galvanometer-based laser synchronization controlling method, calibration method and apparatus, and a LiDAR. The galvanometer-based laser synchronization controlling method includes obtaining a fast-axis feedback signal when a galvanometer scans; obtaining a first phase difference between a fast-axis drive signal and the fast-axis feedback signal and obtaining a second phase difference between an emission period of a laser beam and the fast-axis drive signal; and setting a phase for the fast-axis drive signal based on the first phase difference and the second phase difference.

Abstract: The present disclosure describes membrane compositions and methods for preparing membrane compositions. In particular, the methods employ a layer-by-layer approach to membrane preparation. The membrane compositions provide significantly enhanced membrane performance over existing commercial membranes, particularly in terms of permeability and selectivity.

Abstract: The disclosure relates to an expression method of a Haemocoagulase Acutus (Halase) recombinant protein. The method includes the following steps: (1) optimizing a halase gene; (2) performing Polymerase Chain Reaction (PCR) amplification on an optimized halase gene; (3) constructing an Agkis-pMCX expression vector, transforming plasmids to competent cells of an Escherichia coli for amplification, screening in an Amp-resistant manner for positive cloning, sequencing and extracting recombinant plasmids with correct sequencing; (4) transfecting recombinant plasmids to CHO cells; and (5) expressing the recombinant protein and identifying. According to the expression method, a recombinant halase is expressed first using the CHO cells cultured in a serum-free suspension manner; and by utilizing a bioengineering means, the actual production problems of insufficient raw material sources and unstable quality of a snake venom product are solved successfully.

Abstract: Method of making a cross metathesis product, the method comprising at least step (X) or step (Y): (X) reacting in a cross metathesis reaction a first compound comprising a terminal olefinic group with a second compound comprising a terminal olefinic group, wherein the first and the second compound may be identical or may be different from one another; or (Y) reacting in a ring-closing metathesis reaction two terminal olefinic groups which are comprised in a third compound; wherein the reacting in step (X) or step (Y) is performed in the presence of a ruthenium carbene complex comprising a [Ru?C]-moiety and an internal olefin.

Abstract: A machine learning method predicts whether a specified subject is at high risk of developing cognitive impairment based upon a data record for said specified subject, and a method of training a dual neural network for such. These methods may work by automatically classifying said subject into a first class associated with a first predicted risk of cognitive impairment or a second class associated with a second predicted risk of cognitive impairment. Subject records are obtained comprising a first set of measured data and a second set of data including results for two or more health assessment questionnaires; and a label indicating that the subject belongs to the first class or second class. These records are used to train first and second neural networks, and in the latter method these trained networks are able to predict whether a specified subject should belong to said first class or second class.

Abstract: The present disclosure describes membrane compositions and methods for preparing membrane compositions. In particular, the methods employ a layer-by-layer approach to membrane preparation. The membrane compositions provide significantly enhanced membrane performance over existing commercial membranes, particularly in terms of permeability and selectivity.