Abstract: High resolution object detection systems and methods provide accurate, real-time, one-stage processing, and include a backbone network configured to receive an input image and generate multi-scale feature representations, a feature fusion block configured to fuse the multi-sale feature representations, a plurality of representation transfer modules configured to isolate and decouple sub-task networks and the multi-scale feature representations, and a cascade refinement module configured to process each representation transfer module output to refine predictions. The backbone network generates a plurality of image features corresponding to each of a plurality of image scales and includes a plurality of convolutional layers and a stem block after the first convolutional layer, wherein the stem block improves feature extraction performance. The feature fusion block generates feature outputs for each of a plurality of image scales.

Abstract: A composition includes polydatin and at least one photoisomerization stabilizer and cosmetically acceptable carrier. The at least one photoisomerization stabilizer confers photostabilization to trans-polydatin by at least about 5%, relative to an otherwise identical composition without the at least one photoisomerization stabilizer.

Abstract: Various aspects of the subject technology relate to systems, methods, and machine-readable media for a learning platform that employs machine learning to identify patterns that describe meaningful subgroups of patients associated with a particular disease, distinct characteristics of each subgroup, and implications of belonging to a specific subgroup in terms of developing into a more severe disease state for patient journey mapping. The system may include a mapping server that includes a medical-event embedding engine that embeds electronic health records into vectors, a clustering engine that identifies clusters of patients with similar patient journeys by operating on the output of the medical-event embedding engine, and a cluster profiling engine that identifies distinguishing medical events for each identified cluster by operating on the output of the clustering engine.

Abstract: One or more embodiments of the present specification provide privacy protection-based multicollinearity detection methods, apparatuses, and systems. Data alignment is performed by a member device on respective local feature data with other member devices to construct a joint feature matrix. Privacy protection-based multi-party matrix multiplication computation is performed to compute a product matrix of a transposed matrix of the joint feature matrix and the joint feature matrix. An inverse matrix of the product matrix is determined based on respective submatrices of the product matrix. A variance inflation factor of each attribute feature is determined by the member device with the other member devices using respective submatrices of the inverse matrix and the respective local feature data. Multicollinearity is determined by the member device with the other member devices based on fragment data of the variance inflation factor of each attribute feature.

Abstract: A semiconductor device includes a substrate. An active pattern is on the substrate and extends in a first horizontal direction. First to third nanosheets are sequentially stacked on the active pattern and are spaced apart from each other in a vertical direction. A gate electrode is on the active pattern and extends in a second horizontal direction. The gate electrode surrounds each of the first to third nanosheets. A source/drain region is on the active pattern on at least one side of the gate electrode. An interlayer insulating layer covers the source/drain region. A source/drain contact penetrates the interlayer insulating layer in the vertical direction and is connected to the source/drain region. At least a portion of the interlayer insulating layer is disposed between sidewalls of the source/drain contact and the source/drain region in the first horizontal direction and overlaps sidewalls of the third nanosheet along the first horizontal direction.

Abstract: An aerial vehicle includes a communication unit configured to receive a wireless signal from a geo-fencing device, and a flight controller configured to generate one or more control signals that cause the aerial vehicle to operate in accordance with a set of flight regulations generated based on the wireless signal. The geo-fencing device is configured not for landing of the aerial vehicle. The set of flight regulations includes rules for controlling at least one of the aerial vehicle, a carrier carried by the aerial vehicle, or a payload of the aerial vehicle.

Abstract: A computer-implemented method for treating post-modeling data includes computing, sequentially for each category of a feature, a category importance (CI) value. The CI value is based on a model accuracy change when records of a category being examined are reassigned to a remaining set of categories of the feature according to a cumulative distribution of records among the remaining set of categories of the feature, wherein the remaining set of categories include all categories of the feature, except for the category being examined. A post-modeling category is performed to merge of each category having the CI value less than a CI value threshold.

Abstract: A long-term scheduling method for an industrial byproduct gas system comprises the following steps: dividing information granularity according to the fluctuation features of energy data to form semantic representation of a data sample; with granular data features as input, constructing a deep contrastive network structure through expert scheduling experience data, and constructing knowledge representation under different scheduling states in modes of qualitative and quantitative learning; establishing a fully connected output layer to fit expert scheduling amount, to obtain an initial scheduling policy based on experience knowledge; constructing an actor-critic architecture to calculate a compensation policy that considers long-term scheduling performance.

Abstract: A method and a network device are disclosed for precision time protocol (PTP) clock synchronization. According to an embodiment, a network device determines first estimates of offsets between a slave clock of the network device acting as a PTP slave and a master clock of a PTP master, based on timestamps related to PTP messages communicated between the network device and the PTP master. The network device determines whether there is a change in at least one of a first path from the network device to the PTP master and a second path from the PTP master to the network device. When determining that there is a change in at least one of the first path and the second path, the network device updates the first estimate of the asymmetry based on the first estimates of the offsets.

Abstract: In a method for task execution, a first node receives from a second node a first message that includes first configuration information of a first task. The first node then executes the first task based on the first configuration information.

Abstract: A first node includes a processor. The processor is configured to execute a computer program thereby causing the first node to generate a first message, and send the first message to a third node. The first message is useable to indicate information about a network environment change of a second node. The second node is configured to execute a first task.

Abstract: Volume leveler controller and controlling method are disclosed. In one embodiment, A volume leveler controller includes an audio content classifier for identifying the content type of an audio signal in real time; and an adjusting unit for adjusting a volume leveler in a continuous manner based on the content type as identified. The adjusting unit may configured to positively correlate the dynamic gain of the volume leveler with informative content types of the audio signal, and negatively correlate the dynamic gain of the volume leveler with interfering content types of the audio signal.

Abstract: The present application discloses a method, system, and computer system for identifying dangling records. The method includes obtaining a set of domains, determining whether a record associated with a domain comprised in the set of domains is dangling, and in response to determining that the record associated with the domain is dangling, providing, to a registrant, a notification that the record is dangling.

Abstract: Methods, apparatus and systems for wirelessly transmitting a data flow based on a plurality of automatic repeat request (ARQ) processes are disclosed. In one embodiment, a transmitter module in a wireless communication system is configured for corresponding a data flow to a list of automatic repeat request (ARQ) processes that are associated with a same logical channel, and transmitting data of the data flow based on the list of ARQ processes.

Abstract: This application provides a satellite communication method and apparatus. A terminal obtains a level-1 broadcast signal from a satellite, where a beam status corresponding to the level-1 broadcast signal is an energy saving state, and the level-1 broadcast signal includes synchronization information. The terminal obtains a level-2 broadcast signal from the satellite, where a beam status corresponding to the level-2 broadcast signal is a broadband communication state. The terminal communicates with the satellite based on the level-2 broadcast signal.

Abstract: This application discloses a secure face image transmission method performed by an electronic device. In this application, when a camera component of the electronic device acquires any face image, face image information can be read from a buffer. The face image information is used for indicating a quantity of all face images historically acquired by the camera component. Identification information is embedded in the face image to obtain the face image carrying the identification information. The identification information is used for indicating the face image information but not perceivable by a human being. The face image carrying the identification information is transmitted to a remote server for authenticating the face image using the identification information. Therefore, the security of the face image acquired by the camera component is improved, and the security of the face image transmission process is effectively ensured.

Abstract: A power electronic transformer structure includes: a support, a high-frequency transformer, a base, high-voltage side modules, and a low-voltage side module. The high-voltage side modules are respectively located at the front, top, and back of the support; the low-voltage side module is located at the bottom of the support; the high-frequency transformer is located at the middle of the support.

Abstract: A metal wick material with hierarchical porous structures, which comprises primary pores, secondary pores, and tertiary pores, wherein the pore sizes of the primary pores are greater than those of the secondary pores and the pore sizes of the secondary pores are greater than those of the tertiary pores; the primary pores are obtained by 3D printing, the secondary pores are obtained by volatilizing a binder during a water bath heating process; and the tertiary pores are obtained by forming pores by diffusion and connection among metal powder particles during sintering. A method for fabricating the metal wick material having hierarchical porous structure, comprising: fabricating ordered primary pores having a three-dimensional network structure by a new-type powder extrusion 3D printing process; and then forming secondary pores and tertiary pores inside a macroporous framework by using subsequent treatment processes such as water bath heating and sintering, thus constructing a three-stage pore material.

Abstract: A method for producing sclareol by fermentation of cigar tobacco flower buds is provided, including: mixing cigar tobacco flower buds with deionized water, sterilizing and adding edible yeast, then fermenting at 25-30° C. and 100-300 rpm/min for 24-30 h. This method uses cigar tobacco flower buds as the sole raw material, without adding additional nutrients, and only ferments tobacco flower buds with edible yeast to synthesize sclareol. Meanwhile, this method has extremely high synthesis efficiency.