Abstract: A mineral spring generator has a machine body and a pipeline disposed in the machine body. Two ends of the pipeline respectively have a water inlet and a water outlet. A solenoid valve is connected between the pipeline. At least two injection channels are connected to the pipeline at intervals along a direction of a water flow in the pipeline. A micromotor is connected between each of the at least two injection channels. An outer end of each of the at least two injection channels is connected to a material container. Each material container is adapted to be filled with a liquid mineral spring concentrate. When the present invention is used, water is inputted through the water inlet, and the solenoid valve and the micromotors are activated to add the mineral spring concentrates in the at least two material containers to different positions of the pipeline.

Abstract: A cross-modality knowledge transfer system is provided for adapting one or more source model networks to one or more target model networks. The system is configured to perform steps of providing the TI paired datasets through the source feature encoders of the one or more source model networks, extracting TI source features and TI source moments from the TI paired data by the BN layers of the one or more source model networks, providing the TI paired datasets and the unlabeled TR datasets through the one or more target model networks to extract TI target features and TR target moments, training jointly all the feature encoders of the one or more target model networks by matching the extracted TI target features and TR target moments with the TI source features and TI source moments along with mixing weights, and forming a final target model network by combining the trained one or more target model networks.

Abstract: Systems, methods, and computer-readable media are described for determining the orientation of a target object in an image and iteratively reorienting the target object until an orientation of the target object is within an acceptable threshold of a target orientation. Also described herein are systems, methods, and computer-readable media for verifying that an image contains a target object.

Abstract: Systems are methods are provided for implementing an intelligent optical transceiver. The intelligent optical transceiver implements dynamic health monitoring and “time to fail” prediction functions to predict a failure of a component before it malfunctions during use. By employing the intelligent optical transceiver, a network can prevent failures in its optical connectivity that can degrade the network performance, such as experiencing outages and data unavailability. For example, the intelligent optical transceiver includes a module health monitor, which monitors, in real-time, health parameters of optical communication components in the intelligent optical transceiver. Also, the intelligent optical transceiver includes a time to fail predictor which predicts a time to fail for the optical communication components of the optical transceiver based on a result of a defined regression function.

Abstract: A laterally diffused metal oxide semiconductor structure can include: a base layer; a source region and a drain region located in the base layer; first dielectric layer located on a top surface of the base layer and adjacent to the source region; a voltage withstanding layer located on the top surface of the base layer and located between the first dielectric layer and the drain region; a first conductor at least partially located on the first dielectric layer; and a second conductor at least partially located on the voltage withstanding layer, where the first and second conductors are spatially isolated, and a juncture region of the first dielectric layer and the voltage withstanding layer is covered by one of the first and second conductors.

Abstract: Anomalies in multiple different scenes or images can be detected and localized in a single training flow of a neural network. In various examples, incremental learning can be applied to a given system or network, such that the system or network can learn the distribution of new scenes in a single training flow. Thus, in some cases, when an anomalous image from a new scene is given as input to the network, the network can detect and localize the anomaly.

Abstract: A display substrate includes pixel-defining layer and sub-pixels of N colors, wherein N?3. The pixel-defining layer defines a plurality of sub-pixel areas having the sub-pixels arranged in therein; the pixel-defining layer includes a first pixel-defining layer and a second pixel-defining layer; a height of the second pixel-defining layer is smaller than a height of the first pixel-defining layer; and the first pixel-defining layer is configured to separate adjacent sub-pixels of different colors; the second pixel-defining layer is configured to separate adjacent sub-pixels of the same color; and the sub-pixels of N colors include red sub-pixels, blue sub-pixels, and green sub-pixels; the display substrate includes at least one column of sub-pixels, and in the at least one column of sub-pixels, sub-pixels belonging to a same column have a same color; and in a column of sub-pixels of the same color, corresponding light-emitting layers of at least two adjacent sub-pixels are connected.

Abstract: Embodiments of the present invention disclose a method and system for performing video anomaly detection by training neural networks. The method includes collecting a video of one or more digital images from a source domain. The method includes obtaining a set of images of foreground objects present in the video. In addition, the method includes training of a first neural network to predict frames for the one or more digital images in the video. The first neural network is trained using a future frame prediction module that predicts frames for the one or more digital image. The method includes training of a second neural network to classify the predicted frame as normal and to classify the synthesized pseudo anomaly frame as abnormal. The method includes performing video anomaly detection based on training of the first neural network and the second neural network.

Abstract: A laterally diffused metal oxide semiconductor device can include: a well region having a second doping type; a reduced surface field effect layer of a first doping type formed by an implantation process in a predetermined region of the well region, where a length of the reduced surface field effect layer is less than a length of the well region; a body region of the first doping type extending from a top surface of the well region into the well region; a drain portion of the second doping type extending from the top surface of the well region into the well region; and an insulating structure located between the body region and the drain portion, at least a portion of the insulating structure is located on the top surface of the well region.

Abstract: A laterally diffused metal oxide semiconductor device can include: a base layer; a source region and a drain region located in the base layer; a first dielectric layer located on a top surface of the base layer and adjacent to the source region; a voltage withstanding layer located on the top surface of the base layer and located between the first dielectric layer and the drain region; a first conductor at least partially located on the first dielectric layer; a second conductor at least partially located on the voltage withstanding layer; and a source electrode electrically connected to the source region, where the first and second conductors are spatially isolated, and the source electrode at least covers a space between the first and second conductors.

Abstract: A display substrate includes pixel-defining layer and sub-pixels of three colors. The pixel-defining layer defines a plurality of sub-pixel areas having the sub-pixels arranged in therein; the pixel-defining layer includes a first pixel-defining layer and a second pixel-defining layer; a height of the second pixel-defining layer is smaller than a height of the first pixel-defining layer; and the first pixel-defining layer is configured to separate adjacent sub-pixels of different colors; the second pixel-defining layer is configured to separate adjacent sub-pixels of the same color; and the sub-pixels of three colors include red sub-pixels, blue sub-pixels, and green sub-pixels; the display substrate includes at least one column of sub-pixels, and in the at least one column of sub-pixels, sub-pixels belonging to a same column have a same color; and in a column of sub-pixels of the same color, corresponding light-emitting layers of at least two adjacent sub-pixels are connected.

Abstract: A solid state storage device includes a control circuit, a volatile memory and a non-volatile memory. The non-volatile memory is divided into a first area and a second area. After the host issues a write command and a write data, the control circuit monitors a data amount of the write data continuously stored into the non-volatile memory. Before the data amount of the write data continuously stored into the non-volatile memory reaches a predetermined amount, the write data is stored into plural buffering blocks of the first area in a first write mode. After the data amount of the write data continuously stored into the non-volatile memory reaches the predetermined amount, the write data is stored into plural storing blocks of the second area in a second write mode.

Abstract: This invention provides a method and a system for cleaning a container for storing wafers or masks. A contained with lid is washed directly after the container is loaded with lid opened in the cleaning system. Gas exchange rate in the washing station can be increased such that particles or AMC inside the container or attached to the lid can be carried out more easily. Then, contamination-free gas is purged to the container as well as lid in a high temperature environment. A vacuum station can be optionally adapted between the washing station and the contamination-free gas purging station to enhance the cleanliness of the container.

Abstract: A method of measuring a fuse resistance includes steps as follows. A predetermined voltage value of a force voltage on a common ground (CGND) bus electrically connected to at least one fuse element, a first current value of a measured current through the CGND bus in a first condition, and a second current value of another measured current through the CGND bus in a second condition are preloaded. The second current value is subtracted from the first current value, so as to get a subtracted current value, thereby removing a value of a leakage current through the CGND bus. The predetermined voltage value is divided by the subtracted current value to equal the fuse resistance of the at least one fuse element.

Abstract: Systems, methods, and computer-readable media are described for performing weakly supervised semantic segmentation of input images that utilizes self-guidance on attention maps during training to cause a guided attention inference network (GAIN) to focus attention on an object in an input image in a holistic manner rather than only on the most discriminative parts of the image. The self-guidance is provided jointly by a classification loss function and an attention mining loss function. Extra supervision can also be provided by using a select number pixel-level labeled input images to enhance the semantic segmentation capabilities of the GAIN.

Abstract: A method for enhancing facial/object recognition includes receiving a query image, and providing a database of object images, including images relevant to the query image, each image having a first attribute and a second attribute with each of the first attribute and the second attribute having a first state and a second state. The method also includes creating an augmented database by generating a plurality of artificial images for each image in the database, the artificial images cooperating with the image to define a set of images including every combination of the first attribute and the second attribute in each of the first state and the second state, and comparing the query image to the images in the augmented database to find one or more matches.

Abstract: Aspects include receiving a request to perform an image classification task in a target domain. The image classification task includes identifying a feature in images in the target domain. Classification information related to the feature is transferred from a source domain to the target domain. The transferring includes receiving a plurality of pairs of task-irrelevant images that each includes a task-irrelevant image in the source domain and in the target domain. The task-irrelevant image in the source domain has a fixed correspondence to the task-irrelevant image in the target domain. A target neural network is trained to perform the image classification task in the target domain. The training is based on the plurality of pairs of task-irrelevant images. The image classification task is performed in the target domain and includes applying the target neural network to an image in the target domain and outputting an identified feature.

Abstract: A method for manufacturing a laterally diffused metal oxide semiconductor device and a semiconductor device are provided. A body region is formed before forming a gate dielectric layer and a gate conductor, thereby reducing a channel length of the semiconductor device, thus reducing the on-resistance. In addition, a drift region serves as both a region withstanding a high voltage and a diffusion suppression region for suppressing lateral diffusion of the body region, thereby further reducing the channel length of the semiconductor device, thus manufacturing a short-channel semiconductor device.

Abstract: A method of making a semiconductor structure can include: (i) forming a plurality of oxide layers on a semiconductor substrate; (ii) forming a plurality of conductor layers on the plurality of oxide layers; (iii) forming plurality of thickening layers on the plurality of conductor layers; (iv) patterning the plurality of conductor layers and the plurality of thickening layers to form a hard mask; and (v) implanting ion using the hard mask to form a plurality of doped regions.

Abstract: Systems, methods and devices for real-time contactless elevator service operation of an elevator includes a trained neural network (TNN) model. The TNN model is trained using a training processor with augmented datasets as a synthetic training dataset, to later perform elevator identifier recognition. The augmented data sets are generated from synthetic text images, the synthetic text images are augmented with different geometrical parameters and visual parameters to a predetermined number of variations in appearance to a set of training elevator identifiers. A camera captures a user image. A text image portion from the user image is extracted using the TNN model, and detects an elevator identifier in the extracted text image portion using the extracted text image portion and the TNN model. The detected elevator identifier is displayed for user confirmation or user cancellation, and upon user confirmation, generates a control command based the detected elevator identifier associated with an elevator service.