Abstract: Systems and methods for image segmentation are described. Embodiments of the present disclosure receive an image depicting an object; generate image features for the image by performing a convolutional self-attention operation that outputs a plurality of attention-weighted values for a convolutional kernel applied at a position of a sliding window on the image; and generate label data that identifies the object based on the image features.

Abstract: An image processing system generates an image mask from an image. The image is processed by an object detector to identify a region having an object, and the region is classified based on an object type of the object. A masking pipeline is selected from a number of masking pipelines based on the classification of the region. The region is processed using the masking pipeline to generate a region mask. An image mask for the image is generated using the region mask.

Abstract: An adhesive trap for entrapping pests such as insects such as flying or crawling insects, or other small creatures includes a corrugated panel defining a plurality of channels connected by convex peaks. For example, the corrugated panel, or a plurality of corrugated panels, may be configured and/or assembled to define a pillar or the like. An insect adhesive, such as pressure-sensitive adhesive, is adhered only in the plurality of channels, such that the convex peaks of the corrugated panel are free of adhesive. The adhesive trap is effective at capturing target insects, and the adhesive-free outer portions facilitate handling of the traps without contacting the adhesive and reduces the risk and likelihood of inadvertently ensnaring or adhesively engaging larger, non-target creatures.

Abstract: A polishing tool wear amount prediction device, machine learning device, and system capable of predicting a wear amount of a polishing tool unit of a polishing tool during polishing are provided. The polishing tool wear amount prediction device includes a machine learning device which observes polishing condition data indicating a processing condition of polishing as a state variable indicating a current environment state and performs, based on the state variable, learning or prediction by using a learning model which stores a correlation of the wear amount of the polishing tool with respect to the processing condition of polishing.

Abstract: A method for allocating virtual articles comprises: receiving at least two acquisition requests directed at a virtual article packet, and respectively recording receiving times of the acquisition requests, wherein the virtual article packet comprises one or more virtual article sub-packets; allocating acquisition sequence numbers respectively to the acquisition requests according to a sequential order of the receiving times of the acquisition requests; and allocating the virtual article sub-packets according to the acquisition sequence numbers for the acquisition requests corresponding to the acquisition sequence numbers, and sending virtual articles in the virtual article sub-packets to users corresponding to the acquisition requests.

Abstract: An electronic control actuator of the disclosure includes: a lower housing and an upper housing, the lower housing or the upper housing including a plurality of first matching holes, and the lower housing and the upper housing cooperating to form an installation space; an isolation plate disposed between the lower housing and the upper housing, the isolation plate including a plurality of second matching holes, and a plurality of locating pins respectively passing through the plurality of first matching holes and the plurality of second matching holes; and a compact gear assembly and a drive motor both disposed in the installation space, and the compact gear assembly being cooperatively connected to the drive motor.

Abstract: A semantic understanding method and apparatus, and a device and a storage medium are provided. The method includes: acquiring a recognition character string that matches speech information; acquiring, from an entity vocabulary library, at least one entity vocabulary respectively corresponding to each recognition character in the recognition character string; and according to a situation of each entity vocabulary hitting the recognition character string, determining a matching entity vocabulary as a semantic understanding result of the speech information. By means of the method, insofar as a completely matching entity vocabulary is not acquired, a matching entity vocabulary can still be determined according to an entity vocabulary library, and semantic information of speech is thus accurately understood; and the method also has relatively high fault tolerance for situations such as wrong words, added words, and omitted words, such that the semantic understanding accuracy of speech information is improved.

Abstract: An adhesive trap for entrapping pests such as insects such as flying or crawling insects, or other small creatures includes a corrugated panel defining a plurality of channels connected by convex peaks. For example, the corrugated panel, or a plurality of corrugated panels, may be configured and/or assembled to define a pillar or the like. A pressure sensitive adhesive is adhered only in the plurality of channels, such that the convex peaks of the corrugated panel are free of adhesive. The adhesive trap is effective at capturing target insects, and the adhesive-free outer portions facilitate handling of the traps without contacting the adhesive and reduces the risk and likelihood of inadvertently ensnaring or adhesively engaging larger, non-target creatures.

Abstract: The present disclosure relates to systems, non-transitory computer-readable media, and methods that implement a dual-branched neural network architecture to harmonize composite images. For example, in one or more implementations, the transformer-based harmonization system uses a convolutional branch and a transformer branch to generate a harmonized composite image based on an input composite image and a corresponding segmentation mask. More particularly, the convolutional branch comprises a series of convolutional neural network layers followed by a style normalization layer to extract localized information from the input composite image. Further, the transformer branch comprises a series of transformer neural network layers to extract global information based on different resolutions of the input composite image.

Abstract: Provided are a high quality silicon carbide seed crystal, a silicon carbide crystal, a silicon carbide substrate, and a preparation method therefor. A high quality silicon carbide seed crystal is prepared, the dopant concentrations of a thermal insulation material, a graphite crucible, and a silicon carbide powder material are controlled, a specific crystal growth process and a wafer machining means are integrated, and a high quality silicon carbide substrate is obtained. The obtained silicon carbide substrate has a high crystalline quality and an extremely low amount of micropipes, screw dislocation density, and compound dislocation density; said substrate also has an extremely low p-type dopant concentration, exhibits superior electrical performance, and has a high surface quality.

Abstract: The present disclosure provides a switch circuit and a control method thereof, a smart switch and a control system. The switch circuit includes a wireless communication module, a switch module coupled to a load, a switch-on power obtaining circuit, an energy storage circuit coupled with the wireless communication module and a switch-off power obtaining circuit. The load is coupled with a neutral wire. The switch-on power obtaining circuit is configured to output a first current to the energy storage circuit in response to the switch module being switched on. The switch-off power obtaining circuit is configured to output a second current to the energy storage circuit via the second power supply terminal in response to the switch module being switched off. The second current is less than the first current.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing a plurality of neural networks in a multi-branch pipeline to generate image masks for digital images. Specifically, the disclosed system can classify a digital image as a portrait or a non-portrait image. Based on classifying a portrait image, the disclosed system can utilize separate neural networks to generate a first mask portion for a portion of the digital image including a defined boundary region and a second mask portion for a portion of the digital image including a blended boundary region. The disclosed system can generate the mask portion for the blended boundary region by utilizing a trimap generation neural network to automatically generate a trimap segmentation including the blended boundary region. The disclosed system can then merge the first mask portion and the second mask portion to generate an image mask for the digital image.

Abstract: The present invention discloses a spatial posture adjusting device for an optical axis of a microscopic monitoring system for a cellular fermentation tank. The adjusting device includes a base frame, where the base frame is fixedly provided with an up/down adjustment table; the up/down adjustment table is slidably connected to a load plate; the load plate is fixedly provided with a microscope adjustment mechanism and an illumination rotation adjustment mechanism in sequence; the illumination rotation adjustment mechanism includes a base; the base is fixedly connected to a first support frame; the first support frame is fixedly provided with an illuminator; the microscope adjustment mechanism includes a second support frame; the second support frame is fixedly provided with a clamping device; the clamping device is fixedly provided with a microscope device; the illumination rotation adjustment mechanism is fixedly connected with a mobile platform.

Abstract: The present disclosure relates to systems, non-transitory computer-readable media, and methods that utilize a progressive refinement network to refine alpha mattes generated utilizing a mask-guided matting neural network. In particular, the disclosed systems can use the matting neural network to process a digital image and a coarse guidance mask to generate alpha mattes at discrete neural network layers. In turn, the disclosed systems can use the progressive refinement network to combine alpha mattes and refine areas of uncertainty. For example, the progressive refinement network can combine a core alpha matte corresponding to more certain core regions of a first alpha matte and a boundary alpha matte corresponding to uncertain boundary regions of a second, higher resolution alpha matte. Based on the combination of the core alpha matte and the boundary alpha matte, the disclosed systems can generate a final alpha matte for use in image matting processes.

Abstract: An optical computation system includes a light source configured to produce a pump beam, an optical modulator configured to modulate the pump beam based on the modulation mask to generate a modulated beam, a non-linear medium configured to convert a portion of the modulated beam to a second harmonic (SH) beam and to produce an output including the SH beam and an unconverted portion of the pump beam, and a dichroic mirror configured to receive the output of the non-linear medium and to decouple the SH beam and the unconverted portion of the pump beam, a detector configured to detect a first optical power of the unconverted portion of the pump beam and to detect a second optical power of the SH beam, and a controller configured to generate an updated modulation mask based on the first and second optical powers for transmission to the optical modulator.