Abstract: The present application provides a reworking method of a failed hard mask layer on a via opening in a dielectric layer, including removing the failed hard mask layer; forming an underfill layer to fill the via opening; forming a top hard mask layer on the underfill layer; and forming a mask layer on the top hard mask layer.

Abstract: Low power event driven pixels and reset control circuits for the same are disclosed herein. In one embodiment, an event driven pixel comprises a photosensor; a photocurrent-to-voltage converter coupled to the photosensor; and a difference circuit coupled to the photocurrent-to-voltage converter. The difference circuit includes a source follower transistor and is configured to generate a signal at a gate of the source follower transistor that is based on a voltage output from the photocurrent-to-voltage converter. The difference circuit is further configured to output a difference signal in response to assertion of a row select signal. The event driven pixel can further include a reset control circuit coupled to the difference circuit and configured to initialize the difference circuit, and to reset the difference circuit when the difference signal output from the event driven pixel indicates a change in the voltage greater than a threshold amount.

Abstract: The present application provides an aperture assembly. The aperture assembly is used in an optical lens system. The aperture assembly includes a plate including a hollow portion and a non-hollow portion. The hollow portion is configured to define an aperture of the optical lens system, wherein the contour of the hollow portion has a non-circular shape. The aperture assembly can prevent the size of the aperture from being reduced as the size of the plate is reduced in a case that a circular aperture is applied to the plate. The present application also realizes an aperture adjustment for a non-circular aperture.

Abstract: An embedded system running method includes: allocating, according to a resource dynamic allocation rule, a group of services to be allocated to corresponding operating systems in an embedded system, wherein the embedded system includes a first operating system and a second operating system, and a response speed of the first operating system is higher than a response speed of the second operating system; determining resource allocation results corresponding to the group of services to be allocated, where the resource allocation results are used for indicating, among processing resources of the processor, a processing resource corresponding to each of the group of services to be allocated; and allocating the processing resources of a processor to the first operating system and the second operating system according to an operating system allocation result and the resource allocation result corresponding to each of the group of services to be allocated.

Abstract: Processing methods and apparatus for increasing a reaction chamber batch size. Such a method of processing deposition substrates (e.g., wafers), involves conducting a deposition on a first portion of a batch of deposition wafers in a reaction chamber, conducting an interval conditioning reaction chamber purge to remove defects generated by the wafer processing from the reaction chamber; and following the interval conditioning mid-batch reaction chamber purge, conducting the deposition on another portion of the batch of wafers in the reaction chamber. The interval conditioning reaction chamber purge is conducted prior to exceeding a baseline for acceptable defect (e.g., particle) generation in the chamber and is performed while no wafers are positioned in the reaction chamber.

Abstract: Disclosed are systems and methods for improving interactions with and between computers in content providing, searching and/or hosting systems supported by or configured with devices, servers and/or platforms. The disclosed systems and methods provide a novel recommendation framework that automatically applies transfer learning from a knowledge-base and generating a multi-variable based unified recommendation. The framework dynamically develops a universal word dictionary between a source language and a target language based on Natural Language Processing (NLP word embeddings. This not only leverages the accuracy of multi-lingual product embeddings, but also increases the efficiency and effectiveness in how it can be applied. The disclosed systems and methods, therefore, provides a novel computerized solution for how different knowledges learnt from one electronic platform can be adapted to a global platform for providing global users electronic information.

Abstract: A computer-implemented method includes, for each of a set of training dates: receiving, for each of a sequence of dates including the training date, an input data array representing values of a predetermined set of date-dependent features; receiving a target output corresponding to an evaluation of a predetermined metric at the training date; and performing an update routine including processing the input data array for each date using first layers of a neural network, processing a resulting intermediate data array using second layers of the neural network to generate a network output, and updating values of parameters of the neural network in in a direction of a negative gradient of an error between the target output and the network output. The data processing system is then arranged to generate an embedding array by processing an input data array for each of a given sequence of dates using the first layers of the neural network.

Abstract: The present disclosure provides a display panel and a manufacturing method therefor, and a display apparatus, which relate to the technical field of displaying. The display panel includes a first base plate and a second base plate which are aligned with each other; the first base plate includes a first substrate and a thin film transistor; the thin film transistor includes an active layer; an optical adjustment layer is disposed on the second base plate; an orthographic projection of the optical adjustment layer on the first substrate overlaps with an orthographic projection of the active layer on the first substrate. That is, the optical adjustment layer corresponds to the active layer. In a laminating direction of the display panel, the existence of the optical adjustment layer with a certain height enables a reflecting surface of the second base plate to be closer to the first base plate.

Abstract: The present disclosure is directed to an automated neural architecture search approach for designing new neural network architectures such as, for example, resource-constrained mobile CNN models. In particular, the present disclosure provides systems and methods to perform neural architecture search using a novel factorized hierarchical search space that permits layer diversity throughout the network, thereby striking the right balance between flexibility and search space size. The resulting neural architectures are able to be run relatively faster and using relatively fewer computing resources (e.g., less processing power, less memory usage, less power consumption, etc.), all while remaining competitive with or even exceeding the performance (e.g., accuracy) of current state-of-the-art mobile-optimized models.

Abstract: The present disclosure provides a map information acquisition method and a map acquisition apparatus and a non-transitory computer-readable storage medium. The method comprises: acquiring and constructing, by a map acquisition apparatus, map data, and acquiring, by the map acquisition apparatus, shelf identity information of each of shelves, wherein the shelf identity information is in one-to-one correspondence with the shelves, wherein the map acquisition apparatus comprises: a shopping cart component provided with a map construction system and a radio frequency identification reader; and marking, by the map acquisition apparatus, the map data according to each of the shelf identity information, and generating, by the map acquisition apparatus, map information marked with multiple pieces of shelf identity information.

Abstract: Provided are a method and device for controlling a vehicle and a vehicle, and relates to the technical field of automatic driving.

Abstract: A provider network implements a machine learning deployment service for generating and deploying packages to implement machine learning at connected devices. The service may receive from a client an indication of an inference application, a machine learning framework to be used by the inference application, a machine learning model to be used by the inference application, and an edge device to run the inference application. The service may then generate a package based on the inference application, the machine learning framework, the machine learning model, and a hardware platform of the edge device. To generate the package, the service may optimize the model based on the hardware platform of the edge device and/or the machine learning framework. The service may then deploy the package to the edge device. The edge device then installs the inference application and performs actions based on inference data generated by the machine learning model.

Abstract: A method with a SOTIF scene collection and a self-update mechanism, which is applied to a vehicle, includes a situation judging step, a scene collecting step, a modifying step, a verifying step, and an updating step. As the situation judging step judges that the sensing control dataset belongs to the accident scene dataset, the sensing control dataset collected by the scene collecting step belongs to a SOTIF scene.

Abstract: An example method to facilitate online collaboration includes granting access to a collaboration space to multiple client devices, each associated with a corresponding one of multiple users, the multiple users including a first user and a second user. The method includes receiving input from the first user to create a first set of graphical objects in the collaboration space. The method includes receiving input from the second user to create a second set of graphical objects in the collaboration space. The method includes applying different visual attributes to graphical objects authored by different users to visually identify an author or creator of each graphical object such that all graphical objects of the first set of graphical objects have a first visual attribute and all graphical objects of the second set of graphical objects have a second visual attribute that is different than the first visual attribute.

Abstract: A screen capturing method includes: In a screen split state, an electronic device displays N applications respectively by using N split screens, where one split screen is configured to display an interface of one application, and N is a positive integer greater than 1; further, the electronic device receives a screen capturing operation for a target split screen, where the target split screen includes at least one split screen; finally, in response to the screen capturing operation, the electronic device displays, in full screen, an interface of an application corresponding to each split screen in the target split screen, and generates a screenshot picture, where the screenshot picture includes a full-displayed interface of the application corresponding to each split screen in the target split screen.

Abstract: A display device includes an in-cell touch display panel, a flexible printed circuit board, and a source driver. The in-cell touch display panel includes touch sensing electrodes, touch sensing pins, pixel electrodes, and source pins. The touch sensing pins are electrically connected to the touch sensing electrodes. The source pins are electrically connected to the pixel electrodes. The flexible circuit board includes dummy pads, touch sensing leads, input pads, output pads, and output signal leads. The touch sensing leads are electrically connected to the dummy pads and the touch sensing pins. The output signal leads are electrically connected to the output pads and the source pins. The source driver includes input terminals and output terminals. The input terminals are electrically connected to the input pads. The output terminals are electrically connected to the output pads. The source driver does not include a touch sensing terminal.

Abstract: The present invention relates to an integrated control type hollow curtain, including a front glass panel, a rear glass panel and a curtain body frame. The front glass panel and the rear glass panel are hermetically connected to front and rear surfaces of the curtain body frame to form a hollow curtain with a sealed cavity. An electronic control system and a solar panel are arranged in the top of the sealed cavity. According to the present invention, both the electronic control system and the solar panel are placed in an inner cavity of hollow glass, such that the appearance of a curtain product is simple and attractive, and important components can be protected, thereby avoiding the risk of damage caused by an external environment, and reducing the failure rate.

Abstract: Embodiments of the present disclosure disclose a touch display panel, which includes a plurality of touch sub-units and a plurality of connection wires, wherein the plurality of touch sub-units include the touch sub-units extending in a first direction and a second direction, and two adjacent columns of the touch sub-units extending in the second direction are electrically connected via connection wires; in one of touch electrode repetition units, the connection wires distributed in the first direction are disposed in a misaligned manner in the second direction.

Abstract: In a cylinder deactivation control method of an engine, a cylinder deactivation trigger moment of a to-be-controlled engine is determined; a second cycle and a cylinder deactivation mode corresponding to the next first cycle are determined according to the crankshaft rotation speed at the cylinder deactivation trigger moment, the target torque at the cylinder deactivation trigger moment, and a second cycle and a cylinder deactivation mode corresponding to the previous first cycle at the cylinder deactivation trigger moment; the fuel injection quantity of the ignition cylinder in the next first cycle is determined.

Abstract: Systems, methods, and software described herein manage graph query response using bloom filters and edge indexes. In one implementation, a graph management system may generate indexes and delta indexes for edges in a graph. Once generated, the graph management system may receive a query to the graph and respond to the query using the index and at least one delta index of the delta indexes.