Abstract: A portable blowing device configured for being worn around a neck of a human body, includes two arms each defining an airflow channel therein; and fans received in the arms respectively. The arm includes an inner side wall close to the neck and an outer side wall connected to the inner side wall. The arm includes an air inlet and an air outlet in communication with the airflow channel respectively. The air inlet is arranged at the inner side wall and/or the outer side wall. The fan is configured to generate an airflow passing through the air inlet, the airflow channel and the air outlet in sequence.

Abstract: A speaker comprises a housing, a transducer residing inside the housing, and at least one sound guiding hole located on the housing. The transducer generates vibrations. The vibrations produce a sound wave inside the housing and cause a leaked sound wave spreading outside the housing from a portion of the housing. The at least one sound guiding hole guides the sound wave inside the housing through the at least one sound guiding hole to an outside of the housing. The guided sound wave interferes with the leaked sound wave in a target region. The interference at a specific frequency relates to a distance between the at least one sound guiding hole and the portion of the housing.

Abstract: Disclosed is an integrated packaging method for a portable energy storage device, the method comprising the following steps: 1. preparing a roll cell (1) or a stacked cell (1); and 2. placing the prepared roll cell (1) or stacked cell (1) in a mold, injecting a precursor of an encapsulating agent for carrying out encapsulation, injecting an electrolyte, and completing the encapsulation after the precursor is polymerized.

Abstract: This present disclosure relates to the technical field of artificial intelligence, and provides a vessel image classification method and apparatus, a device, and a storage medium. The method includes: inputting a first vessel image sample into a first image processing model, and obtaining a predicted enhanced image and predicted vessel location information; and training the first image processing model based on a second vessel image sample, vessel location labeling information, the predicted enhanced image, and the predicted vessel location information. In the above solution, the impact of image quality on the vessel classification is considered during training of the vessel classification model, so that an end-to-end vessel classification model subsequently generated based on the trained first image processing model can realize a higher classification accuracy for a low quality vessel image, thereby improving the accuracy of classifying vessels in the vessel image by artificial intelligence.

Abstract: An OLED display panel and fabrication method thereof are provided by the present application. The OLED display panel includes a substrate, a transparent conductive layer, a buffer layer, a metal oxide semiconductor layer, a gate insulating layer, a gate layer, an interlayer insulating layer, a source-drain layer, and an OLED device layer that are stacked. A light emitting direction of the OLED device layer faces the substrate. In the present application, a conductive electrode and a first electrode plate of a transparent capacitor are both patterned and formed with the transparent conductive layer, and only one process is required. Therefore, a mask is saved and fabrication cost is reduced.

Abstract: A SerDes module clock network architecture comprises, a reference clock input port, a plurality of data transmission channels, several user logic interfaces, several frequency division branches and a phase locked loop. The reference lock input port receives an input clock and conveys the input clock to the phase locked loop, the phase locked loop receives the input lock and outputs a PLL output clock signal, the PLL output clock signal is conveyed to the plurality of data transmission channels, and the PLL output clock signal is conveyed to the frequency division branches, and after frequency division, user interface clocks are output and conveyed to the user logic interfaces. When the PLL output clock signal in a SerDes is provided to an internal dedicated channel, several frequency division branches are also divided, and after frequency division, the signal is output to the user logic interfaces for use by an FPGA.

Abstract: A method of driving design on gate electrodes includes steps of: determining position information of a gate-on-array (GOA) device in an available drawing space according to a size design information and a resolution design information of a display panel, based on user configuration; determining target GOA design strategy information used for a current gate electrode driving design among a plurality of preset GOA design strategies; and drawing a GOA device pattern in the available drawing space of the GOA device, based on the target GOA design strategy information.

Abstract: An image attack detection method includes: acquiring an image-to-be-detected, and performing global classification recognition based on the image-to-be-detected to obtain a global classification recognition result; performing local image extraction randomly based on the image-to-be-detected to obtain a target number of local images, the target number being obtained by calculation according to a defensive rate of a reference image corresponding to the image-to-be-detected; performing local classification recognition based on the target number of local images respectively to obtain respective local classification recognition results, and fusing the respective local classification recognition results to obtain a target classification recognition result; and detecting a similarity between the target classification recognition result and the global recognition result, and determining the image-to-be-detected as an attack image when the target classification recognition result and the global classification recognit

Abstract: A light homogenization method for an LCD photocuring 3D printer, including: obtaining the initial grayscale value of a pixel in image data to be printed; obtaining a preset light homogenization compensation value, and calculating the difference between the preset light homogenization compensation value and the initial grayscale value, to obtain a target grayscale value; and controlling the LCD to perform, according to the target grayscale value, photocuring printing on the image data to be printed. The method can solve the problem of uneven exposure in photocuring 3D printing in the prior art, and improve the printing quality. A light homogenization apparatus for an LCD photocuring 3D printer, and a printer are correspondingly provided.