Abstract: Embodiments of the present disclosure disclose a model management apparatus, method and system, where the apparatus includes a model storage module and a model update module; the model storage module is configured to store at least one functional model; the model update module is connected to the model storage module, and is configured to obtain a target functional model to be updated from the model storage module according to a received update instruction, perform functional update on the target functional model to be updated based on a target model data set, and transmit an updated target functional model to the model storage module for storage; where the target functional model is called by a model application apparatus to cause the model application apparatus to implement a set function, and the model application apparatus and the model management apparatus are set up independently of each other.

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: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing a transformer-based encoder-decoder neural network architecture for generating alpha mattes for digital images. Specifically, the disclosed system utilizes a transformer encoder to generate patch-based encodings from a digital image and a trimap segmentation by generating patch encodings for image patches and comparing the patch encodings to areas of the digital image. Additionally, the disclosed system generates modified patch-based encodings utilizing a plurality of neural network layers. The disclosed system also generates an alpha matte for the digital image from the patch-based encodings utilizing a decoder that includes a plurality of upsampling layers connected to a plurality of neural network layers via skip connections.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer-readable media that generate a height map for a digital object portrayed in a digital image and further utilizes the height map to generate a shadow for the digital object. Indeed, in one or more embodiments, the disclosed systems generate (e.g., utilizing a neural network) a height map that indicates the pixels heights for pixels of a digital object portrayed in a digital image. The disclosed systems utilize the pixel heights, along with lighting information for the digital image, to determine how the pixels of the digital image project to create a shadow for the digital object. Further, in some implementations, the disclosed systems utilize the determined shadow projections to generate (e.g., utilizing another neural network) a soft shadow for the digital object. Accordingly, in some cases, the disclosed systems modify the digital image to include the shadow.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for accurately and flexibly generating harmonized digital images utilizing an object-to-object harmonization neural network. For example, the disclosed systems implement, and learn parameters for, an object-to-object harmonization neural network to combine a style code from a reference object with features extracted from a target object. Indeed, the disclosed systems extract a style code from a reference object utilizing a style encoder neural network. In addition, the disclosed systems generate a harmonized target object by applying the style code of the reference object to a target object utilizing an object-to-object harmonization neural network.

Abstract: This disclosure describes one or more implementations of an alpha matting system that utilizes a deep learning model to generate alpha mattes for digital images utilizing an alpha-range classifier function. More specifically, in various implementations, the alpha matting system builds and utilizes an object mask neural network having a decoder that includes an alpha-range classifier to determine classification probabilities for pixels of a digital image with respect to multiple alpha-range classifications. In addition, the alpha matting system can utilize a refinement model to generate the alpha matte from the pixel classification probabilities with respect to the multiple alpha-range classifications.

Abstract: Disclosed is an environment forming apparatus suitable for biological cultivation in extraterrestrial space. The apparatus includes a shell and a biological cabin configured in an upper space of the shell. A light management system for importing light of biological growth from external world is configured in the upper space, a thermal management system for at least balancing a temperature of the upper space and a water supplying system and configured for providing water for creatures growth are both configured on the shell. The biological cabin and the like are set at the appropriate part of the shell, an environment which is relatively suitable for biological growth or cultivation can be created and simulated on an extraterrestrial star, a condition is provided for ecological cultivation of the extraterrestrial space, an ecosystem of the extraterrestrial space can be formed, and a smooth extraterrestrial space biological experiment process is guaranteed.

Abstract: A method of operating an electric power system, which may include: determining a charge state of a first battery unit and a charge state of a second battery unit; determining a difference between the charge state of the first battery unit and the charge state of the second battery unit; and determining whether to enable discharging simultaneously of both the first battery unit and the second battery unit, or to enable discharging of one of the first and second battery units, based on the difference between the charge state of the first battery unit and the charge state of the second battery unit. At least one of the first and second battery units may be a swappable battery unit. The disclosure further relates to an electric power system and to a computer executable code including instructions for operating an electric power system.

Abstract: An electric power system for an electric vehicle, including: first charging and discharging current paths connected to a first battery unit; second charging and discharging current paths connected to a second battery unit; third charging and discharging current paths for a connection with an electric control system of the vehicle; a first power switch between the first discharging current path and the third discharging current path; a second power switch between the second discharging current path and the third discharging current path; a controlling unit to set the first and second power switches based on states of the accumulators, at least one of the first and second battery units being swappable. A controller, a method, and a vehicle comprising the system are also disclosed.

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: A method for controlling a vehicle is provided. The vehicle includes an image capturing device. The method includes: controlling the image capturing device to collect an image of a scene where the vehicle is located; acquiring projection information of a projection pattern in the image; determining image-altering information corresponding to the projection pattern according to the projection information; and controlling movement of the vehicle according to the image-altering information.

Abstract: A method of operating an electric power system, which may include: determining a charge state of a first battery unit and a charge state of a second battery unit; determining a difference between the charge state of the first battery unit and the charge state of the second battery unit; and determining whether to enable discharging simultaneously of both the first battery unit and the second battery unit, or to enable discharging of one of the first and second battery units, based on the difference between the charge state of the first battery unit and the charge state of the second battery unit. At least one of the first and second battery units may be a swappable battery unit. The disclosure further relates to an electric power system and to a computer executable code including instructions for operating an electric power system.

Abstract: The present application discloses a method and apparatus for displaying a virtual landscape picture, a storage medium, and an electronic device. The method includes acquiring a plurality of landscape layer types comprised in the target landscape area to be displayed and determining a type weight corresponding to each landscape layer type in the plurality of landscape layer types; screening at least one target landscape layer type with a corresponding type weight which satisfies a condition from the plurality of landscape layer types, and acquiring an index value of a landscape layer corresponding to each of the screened at least one target landscape layer type; and sequentially storing the index value in each image color channel of the landscape layer index map to obtain the landscape layer index map corresponding to the target landscape area to be displayed.

Abstract: A moulding composition contains a) 75 wt. % to 98.5 wt. % of a polyether block amide, based on a total weight of the moulding composition. The polyether block amide contains a subunit 1, composed of at least one lactam or ?,?-aminocarboxylic acid having 6 to 14 carbon atoms, and a subunit 2, composed of at least one amino- or hydroxy-terminated polyether having at least two carbon atoms per ether oxygen and at least two primary amino or at least two hydroxy groups at the chain ends. The moulding composition also contains b) 1.5 wt. % to 25 wt. % of at least one polyalkenamer, based on a total weight of the moulding composition. The at least one polyalkenamer contains at least one cycloalkene having 5 to 12 carbon atoms. A moulded article can be produced from the moulding composition, which can be a moulding, a film, a bristle, a fibre or a foam.

Abstract: The present disclosure relates to utilizing a neural network having a two-stream encoder architecture to accurately generate composite digital images that realistically portray a foreground object from one digital image against a scene from another digital image. For example, the disclosed systems can utilize a foreground encoder of the neural network to identify features from a foreground image and further utilize a background encoder to identify features from a background image. The disclosed systems can then utilize a decoder to fuse the features together and generate a composite digital image. The disclosed systems can train the neural network utilizing an easy-to-hard data augmentation scheme implemented via self-teaching. The disclosed systems can further incorporate the neural network within an end-to-end framework for automation of the image composition process.

Abstract: Aspects concern a battery pack for mounting on a vehicle. The battery pack may comprise an elongated cell rack comprising one or more battery cells arranged therein. The battery pack may comprise a tubular battery case at least partially surrounding the elongated cell rack. The battery pack may comprise a top seal cover sealing a top portion of the elongated cell rack. The battery pack may comprise a bottom seal cover arranged at a bottom portion of the elongated cell rack. The battery pack may comprise a female pin receptacle arranged on the bottom seal cover at the bottom portion of the elongated cell rack.

Abstract: Provided is a rhomboid structured triboelectric nanogenerator based on a built-in U-shaped support. The device includes a diamond-shaped bracket and two U-shaped supports. The U-shaped supports is connected with the diamond-shaped bracket through four connecting rods, and the supporting plates of U-shaped supports are sequentially adhered with metal conductor layer and a dielectric material layer, which forms the power generation unit. When the diamond-shaped bracket is closed, the dielectric material layers on the two supporting plates come into contact to produce friction charge; and then subjected to the reverse force after compression of the spring, the two dielectric material layers are separated, so as to generate an induction electric field by the friction charge and form a potential difference. Compared with an existing vibration energy harvester, the advantages will be small space occupation, a long service life, simple fabrication, and a high energy conversion rate.

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 in vitro rescue method for immature embryo of avocado includes: collecting immature avocado hybrid fruits, stripping immature zygotic embryos from the sterilized fruits; inducing maturation of zygotic embryos in different media selected according to the maturities of zygotic embryos; and avocado seed embryos growing to normal plants under light conditions after germination. Provided is a technology for the in vitro culture rescue of immature embryo of avocado, wherein the zygotic embryos at different developmental stages are induced to maturation under in vitro conditions, and after maturation, the seed embryos are induced to germinate and form healthy plants. The present invention can effectively improve the survival rate of hybrid avocado zygotic embryo, increase the success rate of hybrid avocado, rescue the hybrid offspring, reduce the loss of hybrid offspring, ensure the genetic integrity of hybrid group, and provide the key technology for hybrid breeding and genetic research of avocado.

Abstract: A power supply on-off detection circuit includes: a signal acquisition circuit, a signal conversion circuit, and a signal processing circuit. The signal acquisition circuit is coupled to a power supply through a switch, and is configured to collect an alternating current signal output by the power supply, and to output the alternating current signal to the signal conversion circuit. The signal conversion circuit is configured to convert the alternating current signal into a voltage signal, and to output the voltage signal to the signal processing circuit. The signal processing circuit is coupled to the signal conversion circuit, and is configured to convert the voltage signal into a level signal. The level signal is used to determine whether the power supply is on or off by determining whether the switch is switched on or off.