Abstract: A present disclosure is a method of segmenting an abnormal robust for complex autonomous driving scenes and a system thereof, specifically relates to the technical field of an image segmenting system. The system includes: a segmentation module, configured to transmit an obtained input image to the segmentation network to obtain a segmentation prediction image, and then quantify the uncertainty of a segmentation prediction by means of calculating two different discrete metrics; a synthesis module, configured to match a generated data distribution with a data distribution of the input image by utilizing a conditional generative adversarial network; a difference module, configured to model and calculate the input image, an generated image, the semantic feature map and the uncertainty feature map based on an encoder, a fusion module and a decoder, to generate the segmentation prediction images for the abnormal objects; a model training module; and an integrated prediction module.

Abstract: The present disclosure provides a carrier aggregation capability reporting apparatus and method, so as to prevent a user equipment (UE) from repeatedly reporting capability information corresponding to a carrier combination supported by the UE, and reduce waste of signaling resources used for reporting. The method includes: determining, by the UE, information about a carrier combination supported by the UE and information about a sub-combination that is of the carrier combination and that is supported by the UE; and reporting, by the UE to a base station, the determined information about the carrier combination supported by the UE and the determined information about the sub-combination that is of the carrier combination and that is supported by the UE.

Abstract: The embodiments of the present disclosure disclose a method and an apparatus for crystal growth. The method for crystal growth may include: placing a seed crystal and a target source material in a growth chamber of an apparatus for crystal growth; executing a growth of a crystal based on the seed crystal and the target source material according to physical vapor transport; determining whether a preset condition is satisfied during the crystal growth process; and in response to determining that the preset condition is satisfied, replacing a sublimated target source material with a candidate source material. In the present disclosure, by replacing the sublimated target source material with the candidate source material, a crystal with large-size and high-quality can be grown.

Abstract: A present disclosure is a method of segmenting an abnormal robust for complex autonomous driving scenes and a system thereof, specifically relates to the technical field of an image segmenting system. The system includes: a segmentation module, configured to transmit an obtained input image to the segmentation network to obtain a segmentation prediction image, and then quantify the uncertainty of a segmentation prediction by means of calculating two different discrete metrics; a synthesis module, configured to match a generated data distribution with a data distribution of the input image by utilizing a conditional generative adversarial network; a difference module, configured to model and calculate the input image, an generated image, the semantic feature map and the uncertainty feature map based on an encoder, a fusion module and a decoder, to generate the segmentation prediction images for the abnormal objects; a model training module; and an integrated prediction module.

Abstract: The present application discloses an integrated structure of MOS transistors having different working voltages. A second spacer of a second MOS transistor having a middle second working voltage is formed by adding a third sub-spacer on the basis of a first spacer of a first MOS transistor having a relatively low first working voltage, and the first spacer is formed by stacking a first sub-spacer and a second sub-spacer. The thickness of the second spacer is adjusted via the third sub-spacer, so as to ensure that a GIDL leakage of the second MOS transistor under the second working voltage satisfies a requirement. The present application also discloses a method for manufacturing an integrated structure of MOS transistors having different working voltages.

Abstract: A method for preventing a network attack, including: receiving, by a first network node in a Ethernet virtual private network (EVPN), a first packet, where the first packet carries a first media access control (MAC) address, and the first MAC address is a source MAC address of the first packet; and determining first MAC entry information, where the first MAC entry information includes a correspondence between the first MAC address, an identifier of the first MAC address, and egress port information of the first MAC address, and the identifier of the first MAC address is used to indicate that an egress port corresponding to the first MAC address is a trusted port. This method can reduce a risk that the EVPN breaks down caused by attacking the EVPN by an attacker.

Abstract: Described are automatically generated software deployment bookmarks that maintain information regarding events and/or triggers with respect to a network software deployment operation. The bookmarks are accessible and useable for evaluating the progress and state of the deployment. For example, a bookmark may be generated when deployment to a client machine succeeds, when a software deployment package is created and/or distributed, and for network component and/or membership state changes. A set of bookmarks may be used to generate a curve indicative of the deployment progress over time, and may be compared to previous progress of deploying other network software to client computer systems that is based on maintained performance history data.

Abstract: Described are automatically generated software deployment bookmarks that maintain information regarding events and/or triggers with respect to a network software deployment operation. The bookmarks are accessible and useable for evaluating the progress and state of the deployment. For example, a bookmark may be generated when deployment to a client machine succeeds, when a software deployment package is created and/or distributed, and for network component and/or membership state changes. A set of bookmarks may be used to generate a curve indicative of the deployment progress over time, and may be compared to previous progress of deploying other network software to client computer systems that is based on maintained performance history data.