31 results about "Visual Manifestations" patented technology

The invention discloses a cloud server performance data collection method and device based on python. According to the sustainably integrated full-automation performance data collection of the cloud server of each node of a cloud provider, the improvement of the performance stability of the cloud service is tested, and performance data at certain time can be automatically obtained and generated; and the visual manifestation of a test result can improve readability, and the execution iteration efficiency of test data in the future is improved. During node testing, a full-automation form and sustainable integration can be implemented, direct manual intervention is not needed, and after data is collected, the collected data is sent to an appointed person and stored locally to serve as the iteration of next data collection. The performance data collection time of cloud service is shortened. The execution iteration efficiency of test data in the future is improved, and the performance indexof the server does not need to be obtained by manual interference in special time periods, such as at night, on weekends and the like.

A video abstract generation method fusing local target features and global features comprises the steps: extracting the local target features of a video, wherein the local target features comprise the visual features of a target, the motion track features of the target and the category label features of the target; constructing a local target feature fusion network by using an attention mechanism, and inputting the local target features to obtain fused local target features; and extracting global features of the video from the video by using an encoder in the encoding-decoding framework, introducing the fused local features into the encoding-decoding framework, fusing global feature information and local target feature information of the video, and obtaining representation vectors with richer expressive force; and decoding a corresponding abstract statement according to the representation vector. According to the method, video local target features are introduced into a video abstract generation model of a coding-decoding framework, the visual expressive force representing the features is enriched, then final text generation is optimized, and related semantic text description is generated based on an input video.

Disclosed is a method for altering the visual performance of a subject using an optical system arranged in front of at least one eye of the subject, and including: b) determining an optotype; c) displaying the optotype to the subject in such a way that it is visible by the subject through the optical system; d) assessing the visual performance of the subject viewing the optotype through the optical system; and e) adding a positive or negative altering value of spherical optical power to the optical system. The method includes, before step b): a) choosing a target value of visual performance; and and in that, at step b), the optotype has a size which is determined depending on the target value of visual performance; at step e), the altering value of spherical optical power is determined depending on the target value of visual performance.

The embodiment of the invention discloses a game synchronization method and device, a terminal, a server and a storage medium. The method comprises the following following steps: constructing an initial game scene , wherein the initial game scene is composed of logic data and a visual representation object; obtaining operation instructions and visual representation objects of other clients from the server; performing synchronization processing on the logic data of the initial game scene according to operation instructions of other clients to obtain synchronized logic data; based on the synchronized logic data, instantiating visual representation objects of other clients in the initial game scene to obtain synchronized scene representation data; and rendering a game picture according to the synchronized logic data and the synchronized scene performance data, so that the game scene is synchronized with other clients. In the embodiment of the invention, the consistency of rendering effects can be improved and the time consumption of game synchronization can be reduced by synchronizing the visual representation object needing instantiation in the game scene. Therefore, the scheme can improve the efficiency of the game synchronization method.

The invention relates to a method for selecting a color filter intended to be worn by a subject in order to improve his visual performance for a given visual task to be realized in a given environment. According to the invention, said method comprises the steps of: —submitting said subject to a colored spatio-temporal visual stimulus adapted to evaluate said visual performance for said given visual task, said stimulus being designed so that its chromaticity may be modified; —evaluating the visual performance of the subject based on his visual perception of said visual stimulus, the chromaticity of the visual stimulus taking at least two distinct values, each chromaticity being uniform; and —selecting said color filter based on the results of said evaluation. The invention further proposes an optical article comprising such a color filter. The invention finally provides a system for selecting a color filter using said method.

The invention relates to a material display method and device, electronic equipment, a storage medium and a program product, and the method comprises the steps: obtaining a shot video image, obtaining a track image based on the position of a target recognition object in the video image, and carrying out the superposition processing according to the track image and a material image corresponding to a target element, obtaining a plurality of target unit special effect images; and combining the plurality of target unit special effect images to obtain a target element, the main body of the target three-dimensional special effect is a solid geometric structure comprising a plurality of surfaces, and the plurality of target unit special effect images are respectively attached to the corresponding geometric surfaces so as to display the plurality of target unit special effect images on the plurality of geometric surfaces of the solid geometric structure. The target element is generated based on the action of the user, so that the user interaction experience can be improved; and the main body of the target element in the generated special effect is a three-dimensional structure, so that the visual expressive force of the special effect can be greatly improved.

A method for detecting false information based on a common-attention multi-modal fusion mechanism comprises the following steps: S1, constructing a BERT model for extracting text features; s2, constructing an R-CNN model for extracting visual features; s3, constructing a co-attention converter layer Co-TRM model fusing the text and the visual features; s4, inputting the fused feature vectors into a domain classifier, mapping multi-modal features of different information to the same feature space, classifying texts into different events, and deleting special features of the events; and S5, inputting the fused feature vectors into a false information detector, and judging whether the information is true or false by using potential multi-modal features. According to the method, a common attention mechanism is utilized, a random pixel sampling mechanism is used for enhancing the robustness of visual performance, improving a multi-modal model structure and respectively enhancing a text feature extraction method and a visual feature extraction method, so that information exchange among different modals is improved, and the accuracy of false information detection is improved.

The invention belongs to the technical field of image processing, and particularly relates to a target edge extraction method and an image segmentation method and system. According to the target edge extraction method, an unsupervised rough segmentation method based on edge search is constructed to simulate a visual working memory principle, overall cognition of a scene is carried out, and an accurate edge of a large-area target is obtained. The image segmentation method is based on a target extraction method and a segmentation neural network, visual attention is simulated to carefully observe a scene, and small-area details are obtained. And finally, generating a segmentation result by combining an accurate edge in rough segmentation according to the inference ability shown visually. According to the method, on the basis of simulating visual reasoning to respectively recognize the whole and details of the ground scene, mutual complementation and mutual correction of the whole information and the detail information are carried out, noise removal and error repair are implemented, and thus the image segmentation precision is effectively improved.

A system and method is provided for attaining sustainable productivity growth in an entity. The system and method provides visual representations of projected entity valuations, and productivity growth based on selected areas of improvement, including improved gross margins in manufacturing entities. The visual representations and projected values can be provided on a graphical user interface, that can be adjusted by the user.