Abstract: An Artificial Intelligence (AI) driven pull request summarization system utilizes a large language model to classify the code changes of a pull request and to generate a summary of the changes contained in the pull request. The large language model predicts the pull request classification and summarization given a prompt that includes the top-k code changes in the pull request of a repository. The code changes are prioritized based on the most impact that a code change has on the files, methods and classes in the repository. Each of the top-k code changes is linked to a related open issue of the repository, if any. A suggested code reviewer for a code change is then selected from an author or commentator associated with the linked open issue.

Abstract: The disclosure provides a security analysis method and system based on protocol state, which relates to the technical field of protocol security protection. The method includes the following: a node traversal table is built, the node traversal table is scanned and analyzed according to the protocol trigger sequence rule, a first security evaluation factor of a protocol stack is determined, and a second security evaluation factor of each protocol is determined based on protocol normal application rule, and the trustworthiness degree of the second security factor is determined based on the first security factor, and the second security factor is revised based on the trustworthiness degree, and the security state of the protocol is determined according to the revised second security factor, thus the analysis of the protocol state is realized, and the security of the protocol can be accurately determined.

Abstract: Disclosed is a pen for planting spherical beads. A hollow hole is formed in a lower end of a pen body. A jacking rod vertically extending in a downward direction and a vertical sliding component are under the pen body. A cavity is formed inside the vertical sliding component, a bead inlet hole is formed on the top of the cavity, a bead outlet hole is formed in the bottom of the cavity, the bottom of the vertical sliding component is inclined, and the bead outlet hole is positioned at the lowest end of the cavity of the vertical sliding component. When the vertical sliding component downwards slides to the lowest point of a track relative to the pen body, a vertical distance between the lower end surface of the jacking rod and the bead outlet hole can accommodate a spherical bead.

Abstract: The present disclosure provides a A driving method of a display panel, including: determining a target region and a non-target region in the display panel; dividing the target region into a plurality of target sub-regions, and dividing the non-target region into a plurality of non-target sub-regions, wherein an area of the target sub-region is less than that of the non-target sub-region; determining brightness compensation values of each target sub-region and each non-target sub-region according to a gray scale value of each pixel in an input image; determining a compensated brightness of each pixel unit in the display panel according to the gray scale value of each pixel in the input image, the brightness compensation values of each target sub-region and each non-target sub-region; and driving each pixel unit in the display panel to emit light according to the compensated brightness of each pixel unit in the display panel.

Abstract: A speech recognition model includes an encoder network, a prediction network, and a joint network. The encoder network is configured to receive a sequence of acoustic frames characterizing an input utterance; and generate, at each of a plurality of output steps, a higher order feature representation for a corresponding acoustic frame in the sequence of acoustic frames. The prediction network is configured to: receive a sequence of non-blank symbols output by a final Softmax layer; and generate, at each of the plurality of output steps, a dense representation. The joint network is configured to generate, at each of the plurality of output steps based on the higher order feature representation and the dense representation, a probability distribution over possible speech recognition hypotheses. The joint network includes a stack of gating and bilinear pooling to fuse the dense representation and the higher order feature representation.

Abstract: Aspects of the present disclosure provide techniques, including devices and systems implementing the techniques, to manage ambient noise in a wearable audio output device to facilitate increased awareness for a user of the wearable audio output device. One example technique for managing ambient noise generally includes determining a first event, ducking an audio level of the wearable audio output device from a first level to a second level based on the determination, monitoring for a second event, and ducking the audio level of the wearable audio output device to a third level based on the monitoring, wherein the second level is different than the third level. Such techniques may help to more accurately determine the occurrence of events that are important to the user and manage the wearable audio output device to facilitate user awareness, as well as mitigate the undesirable consequences of events that are unimportant to the user.

Abstract: A driving method for a display device, including: inputting a source image, wherein the source image includes depth information; searching in the source image by utilizing the depth information to determine first image pixel points matched with light-emitting sub-pixels of the display device; assigning pixel gray values of the first image pixel points to the light-emitting sub-pixels of the display device; and controlling the light-emitting sub-pixels to emit light according to assigned pixel gray values, so that the display device displays a multi-viewpoint image corresponding to the source image. The driving method of the display device provided in the embodiment of the present application directly searches for the first image pixel that satisfies the preset parallax condition on the source image based on the depth information, renders the display device, saves hardware storage resources, and improves the processing efficiency of multi-viewpoint naked-eye 3D display.

Abstract: A dimming method for a display panel is provided, including: acquiring image data of an image to be displayed; determining initial dimming pixel data of each of dimming pixels according to the image data; determining a dimming pixel to be filtered in the dimming screen according to the initial dimming pixel data of each of dimming pixels; determining a first filtering region according to the determined dimming pixel to be filtered; determining a filtering parameter of each of dimming pixels in the first filtering region according to a relationship between each of dimming pixels in the first filtering region and initial dimming pixel data of each of dimming pixels in the first filtering region; and updating dimming pixel data of the dimming pixel to be filtered according to the filtering parameter of each of dimming pixels in the first filtering region, so as to obtain updated dimming pixel data.

Abstract: An image processing method is applied to an image display device, the image display device includes a lens having distortion coefficients and a display screen. The image processing method includes: dividing a display region of the display screen into image regions according to the distortion coefficients of the lens, an outer boundary line of each image region enclosing a polygon, a geometric center of the polygon enclosed by the outer boundary line of the image region coinciding with a geometric center of the lens, and distortion coefficients of positions of the lens on which vertexes of the polygon enclosed by the outer boundary line of the image region are mapped being the same; and performing anti-distortion processing on coordinates of vertexes of the image region according to a distortion coefficient corresponding to the vertexes of the image region to obtain texture coordinates of the vertexes of the image region.

Abstract: A modulation classification and demodulation system is disclosed having a modulation classifier and a demodulator to classify modulations and then demodulate received signals that deviate during propagation from noise. The modulation classifier classifies each type of modulation for each signal transmitted to the receiver in a sequenced modulation signal into a corresponding modulation class based on a modulation classifier neural network that identifies each type of modulation. The sequenced modulation signal that is received by the modulation classifier is deviated from when initially transmitted by the transmitter based on noise that impacts the sequenced modulation signal as the sequenced modulation signal propagates from the transmitter. The demodulator demodulates the receives signal by decoding each symbol included in the received signal based on the determined modulation class of the received signal.

Abstract: A sound-generating device, a display device, a sound-generating controlling method, and a sound-generating controlling device are provided. The sound-generating device includes: a reflection plate which includes a first sound wave reflection face arranged towards a first direction; a plurality of main loudspeakers, the plurality of main loudspeakers are distributed in an array in a preset three-dimensional space, and the preset three-dimensional space is located at one side of the first sound wave reflection face towards a first direction; the plurality of main loudspeakers include first main loudspeakers with a sound-generating direction towards the first direction, and second main loudspeakers with a sound-generating direction towards a second direction, the second direction is an opposite direction of the first direction; and the sound waves emitted by the second main loudspeakers are transmitted to the first sound wave reflection face and can be reflected by the first sound wave reflection face.

Abstract: A maglev vehicle and a suspension frame assembly thereof. The suspension frame assembly includes multiple suspension frames sequentially connected to each other. Each suspension frame comprises two motor beams, four support arms, and four air springs. An air spring mounting seat is disposed at a top portion of each support arm. The air spring mounting seats are cavities having openings. The air springs are accommodated in the respective cavities.

Abstract: A display method comprises: acquiring multiple frames of images to be displayed, one frame of two adjacent frames of the images to be displayed being an image in an odd-numbered row, the other frame being an image in an even-numbered row, the resolution of the image in the odd-numbered row being M*(a*N), the resolution of the image in the even-numbered row being M*(a*N) or (M?1)*(a*N), and a being 1 or 2; and causing an odd-numbered display group to display the image in the odd-numbered row, and causing an even-numbered display group to display the image in the even-numbered row, the odd-numbered display group comprising M*(a*N) odd-numbered display units, the even-numbered display group comprising (M?1)*(a*N) even-numbered display units, the odd-numbered display unit in the i-th row comprising a first sub-pixel, a second sub-pixel, and a third sub-pixel in a pixel unit in the i-th row.

Abstract: Provided in the present application are an anti-periostin humanized monoclonal antibody, and a preparation method therefor and the use thereof. The anti-periostin humanized monoclonal antibody has a relatively high expression level, and also has a better binding specificity and affinity to an antigen.

Abstract: Certain aspects provide methods and apparatus for recovering audio quality of voice when processing signals associated with a wearable audio output device. A method that may be performed includes receiving. by an in-ear microphone acoustically coupled to an environment inside an car canal of a user, an audio signal having a first frequency band. predicting high-frequency band information for the audio signal using a model trained using training data of known high-frequency bands associated with low-frequency bands. generating an output signal having a second frequency band based. at least in part. on the first frequency band of the audio signal and the predicted high-frequency band information for the audio signal, and outputting. by the wearable audio output device. the output signal having the second frequency band.

Abstract: A liquid crystal display device, a method for displaying an image, and an electronic device are provided. In the first operation mode, a control circuit of the display device is configured to: receive display data of an mth image frame, and control the display device to display n sub-field images. The display data includes initial gray scales under n base colors. In a sub-field image of the mth image frame: pixels include a target pixel, and an actual gray scale of the target pixel is an average of an intermediate gray scale and initial gray scales of T non-target pixels. An actual gray scale of at least a part of non-target pixels is an initial gray scale thereof. The target pixel meets at least one of a first condition and a second condition.

Abstract: A method for preparing a sensing device is provided, the method includes: forming, on a surface of a transparent substrate, a first auxiliary patterning layer having a trench; and forming a mesh structure in the trench of the first auxiliary patterning layer by means of an electroplating process, or forming a conductive layer in the trench of the first auxiliary patterning layer by means of an electroplating process and then etching the conductive layer to form a mesh structure. The mesh structure has a line width of less than or equal to 1.5 microns and a thickness of greater than or equal to 2 microns.

Abstract: A display device, a sound producing control method, a method and device for determining parameters. The display device includes a plurality of exciters arranged on the display panel, the display panel producing sound on a screen through the plurality of exciters; wherein, the plurality of the exciters are formed as at least one sound producing unit, each sound producing unit includes at least two exciters, and each exciter in the same sound producing unit is connected to a same audio signal input end, in one sound producing unit, a signal filter is provided on an audio transmission line from the audio signal input end to at least one exciter, and used for low frequency compensation of a sound produced by the entire sound producing unit.

Abstract: A path planning method and a biped robot using the same are provided. The method includes: generating a candidate node set for a next foot placement based on a biped robot's own parameters and joint information of a current node, adding valid candidate nodes in the candidate node set to a priority queue so as to select optimal nodes for realizing next node expansion. These optimal nodes are output to generate a foot placement sequence from an initial node to a target node, which can greatly reduce the search amount for path nodes when the robot's legs intersect and touch the ground, thereby improving the efficiency of path planning.

Abstract: Disclosed is a backlight diffusion parameter generation method, including: selecting and measuring illumination diffusion data of at least one of multiple light emitting regions, wherein the illumination diffusion data includes brightness data of multiple pixels on a display panel and distance data between positions corresponding to multiple pixels and a position where an illuminated light emitting region is located when only one light emitting region is illuminated; preprocessing the illumination diffusion data to obtain an effective pixel; performing function fitting according to data corresponding to multiple effective pixels to obtain a point spread function representing a relationship between diffusion brightness and a diffusion distance; storing a backlight diffusion parameter lookup table, which includes multiple diffusion distance bonding points and diffusion brightness corresponding to each of the diffusion distance bonding points, and also includes a diffusion slope between two adjacent diffusion di