Abstract: A hyperspectral remote sensing image classification method based on a self-attention context network is provided. The method constructs a spatial dependency between pixels in a hyperspectral remote sensing image by self-attention learning and context encoding, and learns global context features. For adversarial attacks in the hyperspectral remote sensing data, the proposed method has higher security and reliability to better meet the requirements of safe, reliable, and high-precision object recognition in Earth observation.

Abstract: Optical communication equipment used in data center interconnection (DCI) applications includes a chassis, having one or more panel modules or line cards and an earring bracket for securing the chassis. Also included is an extended management bracket, extending from the chassis, and an extended function module extending from the extended management bracket, located in an area in front of the chassis. A cable can be provided between the chassis and the extended function module, supporting data transmission therebetween. Alternatively, or in addition, a printed circuit board assembly can be mounted within the chassis, extending outside the chassis, along or within the extended management bracket, from a proximal end to a distal end of the extended management bracket, configured for connection to components of the extended function module, where the printed circuit board assembly supports data transmission and/or power supply between the chassis and components of the extended function module.

Abstract: A hyperspectral remote sensing image classification method based on a self-attention context network is provided. The method constructs a spatial dependency between pixels in a hyperspectral remote sensing image by self-attention learning and context encoding, and learns global context features. For adversarial attacks in the hyperspectral remote sensing data, the proposed method has higher security and reliability to better meet the requirements of safe, reliable, and high-precision object recognition in Earth observation.

Abstract: The disclosure relates to a prefabricated movable photosensitively adjustable solar bionic landscape pavilion capable of being disassembled and constructed quickly. The landscape pavilion comprises a support structure, a ceiling, a movable platform, a linkage system, a lighting system, and an energy system. The support structure is arranged at the center of the movable platform and forms large-area shelter space with the ceiling. The ceiling may rotate around the support structure in the height direction of the support structure, so as to be unfolded or folded. The movable platform is a base of the landscape pavilion, undertakes the support structure and the ceiling above the movable platform, and meanwhile, provides fixing space for the linkage system, the lighting system, and the energy system.

Abstract: Disclosed are a method for updating a neural network and an electronic device. The method includes: inputting a first image set having tag information into a first depth neural network, and determining a cross entropy loss value of the first image set by using the first depth neural network; inputting a second image set having no tag information separately into the first depth neural network and a second depth neural network, and determining a consistency loss value of the second image set, the first depth neural network and the second depth neural network having the same network structure; updating parameters of the first depth neural network based on the cross entropy loss value and the consistency loss value; and updating parameters of the second depth neural network based on the updated parameters of the first depth neural network.

Abstract: The disclosure relates to a prefabricated movable photosensitively adjustable solar bionic landscape pavilion capable of being disassembled and constructed quickly. The landscape pavilion comprises a support structure, a ceiling, a movable platform, a linkage system, a lighting system, and an energy system. The support structure is arranged at the center of the movable platforms and forms large-area shelter space with the ceiling. The ceiling may rotate around the support structure in the height direction of the support structure, so as to be unfolded or folded. The movable platform is a base of the landscape pavilion, undertakes the support structure and the ceiling above the movable platform, and meanwhile, provides fixing space for the linkage system, the lighting system, and the energy system.

Abstract: Disclosed are a method for updating a neural network and an electronic device. The method includes: inputting a first image set having tag information into a first depth neural network, and determining a cross entropy loss value of the first image set by using the first depth neural network; inputting a second image set having no tag information separately into the first depth neural network and a second depth neural network, and determining a consistency loss value of the second image set, the first depth neural network and the second depth neural network having the same network structure; updating parameters of the first depth neural network based on the cross entropy loss value and the consistency loss value; and updating parameters of the second depth neural network based on the updated parameters of the first depth neural network.