Abstract: A vacuum switching valve and a suction system having the same. The vacuum switching valve comprises: a valve body, comprising a first end and a second end, the second end being provided with an air inlet, an air outlet and a through hole; a valve element movably arranged in the valve body; a cylinder, the cylinder being connected to the first end and the valve element, the cylinder drives the valve element to move in the valve body, to close or open the air inlet; a stopper passing through the through hole, the stopper comprising a third end and a fourth end, the third end being connected to the valve element, the fourth end being located on the side of the through hole away from the valve element.

Abstract: A system for identifying movers in an independent cart system includes movers having at least one magnet and sensors generating a feedback signal responsive to detecting a magnetic field from the magnet as each mover travels past the sensor. A memory stores an identifier and a corresponding digital fingerprint for each mover. The stored digital fingerprint is generated as a function of the magnetic field generated by the magnet on each mover. A controller receives the feedback signal from each sensor and determines a run-time digital fingerprint for each mover corresponding to the magnetic field generated by the magnet on each mover as a function of the feedback signal. The run-time digital fingerprint is matched to one of the stored digital fingerprints, and the identifier, corresponding to the stored digital fingerprint matching the run-time fingerprint, is read from memory.

Abstract: An image processing method and apparatus may be provided. The process may include, obtaining source compressed texture data of a target image by encoding the target image by using a source compressed texture format and determining a target compressed texture format adapted to the display card and a target compressed block size corresponding to the target compressed texture format. The process may also include obtaining a plurality of pieces of image texture data, based on decoding and aligning the source compressed texture data by using the target compressed block size, and obtaining target compressed texture data of the target image, based on transcoding each piece of the plurality of pieces of image texture data by using the target compressed texture format and the target compressed block size. The process may also include rendering the target image based on the target compressed texture data.

Abstract: Active semiconductor devices in an integrated circuit are provided lateral electrical isolation by surrounding narrow deep trench isolation regions that are merged at shared portions of the narrow deep trench isolation regions. A wide deep trench isolation region laterally surrounds the merged narrow deep trench isolation regions.

Abstract: Provided are computer-implemented methods for generating embeddings for objects which may include receiving heterogeneous network data associated with a plurality of objects in a heterogeneous network; selecting at least one pattern of objects; determining instances of each pattern of objects based on the heterogeneous network data; generating a pattern matrix for each pattern of objects based on the instances of the pattern of objects; generating pattern sequence data associated with a portion of each pattern matrix; generating network sequence data associated with a portion of the heterogeneous network data; and combining the pattern sequence data and the network sequence data into combined sequence data. In some non-limiting embodiments or aspects, methods may include generating a vector for each object of the plurality of objects based on the combined sequence data. Systems and computer program products are also provided.

Abstract: Provided are methods for enhancing a distribution of graph feature embeddings in an embedding space to improve discrimination of graph features by a graph neural network (GNN) that may include receiving a dataset comprising graph data associated with a graph, calculating a distance between a first set of node embeddings and a second set of node embeddings, determining a measure of uniformity for the dataset, determining a plurality of groups of node embeddings, determining a measure of alignment for the plurality of groups of node embeddings, generating a set of graph features based on the measure of uniformity, the measure of alignment, and the distance, and training the GNN based on the set of graph features to provide a trained GNN. Systems and computer program products are also disclosed.

Abstract: A speaker includes a frame, a magnetic circuit system with a magnetic gap, and a vibration system, the vibration system includes a diaphragm assembly, a voice coil assembly, a first elastic wave assembly fixed, and a second elastic wave assembly, the diaphragm assembly includes a dome, a suspension, and a connecting wall extending from the dome to the voice coil assembly, the connecting wall includes a connecting portion connected with the dome the suspension, and a first suspension portion and a second suspension portion located at a side away from the dome, one end of the first elastic wave assembly away from the frame is connected with the first suspension portion, one end of the second elastic wave assembly away from the frame is connected with the second suspension portion. Compared with the related art, the speaker disclosed by the present disclosure could improve space utilization.

Abstract: A developing cartridge includes a casing configured to accommodate developer and having a first side and a second side oppositely arranged in a first direction, and a third side and a fourth side oppositely arranged in the second direction, the first direction and the second direction being intersected with each other. The developing cartridge further includes a developing roller, positioned at the third side, for rotating around a first axis extending in the first direction; a coupling, positioned at the first side, for receiving driving-force output from an image forming device to rotate; a driving surface, inclined to the first direction, including a first driving surface and a second driving surface; a detected part, at least partially positioned at the second side, for moving according to rotation of the coupling.

Abstract: The disclosure relates to the technical field of static synchronous reactive power compensation devices, provides an updating and optimization system for a static synchronous reactive power compensation device. The first module sets multiple contamination detection points on the static synchronous reactive power compensation device, and constructs multiple updating and optimization contamination data chains according to contamination data. The second module constructs a related contamination data set according to the related contamination data, and calculates the sub-update optimization influence coefficients. The third module analyzes and calculates the remaining related pollution data, and constructs sub-update optimization influence coefficient sets. The fourth module sorts the sub-update optimization influence coefficient sets, determines the comprehensive updating optimization influence coefficient set, and calculates the comprehensive updating optimization influence coefficient.

Abstract: System, methods, and other embodiments described herein relate to implementing cooperative management strategies. In one embodiment, a method includes determining a severity score and a risk area. The method may then determine whether vehicles should belong to a first vehicle set, which may perform a cooperative vehicle response action, or a second vehicle set, in which individual response actions may be performed. The method may then generate a cooperative response action or an individual response action corresponding to the cooperative or individual vehicle sets, provided such vehicle sets are not empty.

Abstract: Methods, systems, and computer program products for universal depth graph neural networks may obtain a graph G including an adjacency matrix A and an attribute matrix X and train a graph convolutional network using according to the following updating Equation: H=?(?dXW), where H is an output embedding matrix, ?(·) is a nonlinear activation function, ? is a filter, X is an attribute matrix, W is a trainable weight matrix, and d is a trainable depth parameter.

Abstract: Methods, systems, and computer program products for knowledge graph based embedding, explainability, and/or multi-task learning may connect task-specific inductive models with knowledge graph completion and enrichment processes.

Abstract: Described are a method, system, and computer program product for generating robust graph neural networks using universal adversarial training. The method includes receiving a graph neural network (GNN) model and a bipartite graph including an adjacency matrix, initializing model parameters of the GNN model, initializing perturbation parameters, and sampling a subgraph of a complementary graph based on the bipartite graph. The method further includes repeating until convergence of the model parameters: drawing a random variable from a uniform distribution; generating a universal perturbation matrix based on the subgraph, the random variable, and the perturbation parameters; determining Bayesian Personalized Ranking (BPR) loss by inputting the bipartite graph and the universal perturbation matrix to the GNN model; updating the perturbation parameters based on stochastic gradient ascent; and updating the model parameters based on stochastic gradient descent.

Abstract: A method and system for determining an access score is disclosed. The method includes receiving an access request to access a resource by a user device. Next, a user embedding is retrieved from an embedding table, the user embedding associated with a user identifier of the user device and providing a multidimensional data point that represents a context of a user identifier. The context may correspond to the user identifier appearing in previous access requests within temporal proximity to other access requests from a subset of other user devices among a plurality of user devices. The method then inputs the user embedding into a first machine learning model that is trained based at least in part on the embedding table. The first machine learning model subsequently outputs an access score that corresponds to a level of authenticity of authorizing the user device to access the resource.

Abstract: Provided in the present disclosure are a method and an apparatus for controlling a source driver, and a display system.

Abstract: A system for identifying movers in an independent cart system includes movers having at least one magnet and sensors generating a feedback signal responsive to detecting a magnetic field from the magnet as each mover travels past the sensor. A memory stores an identifier and a corresponding digital fingerprint for each mover. The stored digital fingerprint is generated as a function of the magnetic field generated by the magnet on each mover. A controller receives the feedback signal from each sensor and determines a run-time digital fingerprint for each mover corresponding to the magnetic field generated by the magnet on each mover as a function of the feedback signal. The run-time digital fingerprint is matched to one of the stored digital fingerprints, and the identifier, corresponding to the stored digital fingerprint matching the run-time fingerprint, is read from memory.

Abstract: A system and method is provided for implementing computer-controlled systems (e.g., vehicle safety and traffic management systems) based on compliance rates. A vehicle system and method are provided that estimate compliance rates in real-time, and have the capability to dynamically modify the computer-controlled instructions generated by a computer-controlled system (e.g., vehicle safety and traffic management systems) in a manner that improves compliance, and in-turn improves overall system performance. For example, a vehicle includes sensors receiving data from a plurality of communicative connected vehicles. The received data is associated with a real-time monitoring of traffic conditions for the vehicle. The vehicle also includes a controller device estimating a compliance rate in real-time based on the received data. The controller device of the vehicle also controls an automated operation of the vehicle using a modified computer-controlled instruction based on the estimated compliance rate.

Abstract: A method performed by a computer is disclosed. The method includes receiving a dataset including one or more data fields. The computer can then receive instructions to compute an aggregate feature of the dataset. The computer can then determine an operator and a window based on the aggregate feature to be computed. The method may then include the computer processing the dataset. The compute may then apply the operator with the window to the processed dataset to compute the aggregate feature of the dataset.

Abstract: Systems, methods, and computer program products are provided for saving memory during training of knowledge graph neural networks. The method includes receiving a training dataset including a first set of knowledge graph embeddings associated with a plurality of entities for a first layer of a knowledge graph, inputting the training dataset into a knowledge graph neural network to generate at least one further set of knowledge graph embeddings associated with the plurality of entities for at least one further layer of the knowledge graph, quantizing the at least one further set of knowledge graph embeddings to provide at least one set of quantized knowledge graph embeddings, storing the at least one set of quantized knowledge graph embeddings in a memory, and dequantizing the at least one set of quantized knowledge graph embeddings to provide at least one set of dequantized knowledge graph embeddings.

Abstract: A navigation method, device, and system are provided in this application. The method includes: obtaining a first target location and a second target location (101), where the first target location is a location closest to a start location on K road segments, the second target location is a location closest to an end location on the K road segments, and K is a positive integer; obtaining a first path from the first target location to the second target location based on a vector pathfinding algorithm and the K road segments (102); obtaining a second path from the start location to the first target location and a third path from the second target location to the end location (103); and determining a fourth path from the start location to the end location (104).