7results about How to "Accurate labeling" patented technology

Embodiments of the present specification provide a sample processing method and device, wherein the sample processing method comprises: obtaining a to-be-processed sentence; constructing a word group corresponding to the to-be-processed sentence, and determining a feature label corresponding to each word unit in the word group; constructing at least one negative sample and a target positive sample according to the feature label and the word group; calculating a sample similarity between the target positive sample and each negative sample, and determining a target negative sample in the at least one negative sample according to the sample similarity.

The application provides a kind of geographical information intelligent analysis system and method for land space planning, obtains planning base map by receiving multi-source planning data;The planning rules associated with each geographical entity in the planning base map are deconstructed, the main chain path and the branch path are identified, and the spatial access conditions of each land use type are extracted;Based on spatial overlay analysis, the spatial conflicts among the three zones and three lines of each planning layer are detected, the land use control conflicts are determined according to the spatial access conditions, and the time conflicts are detected;Each conflict detection result is associated with the corresponding triggered branch path, the conflicts of the corresponding main chain path are corrected, and the comprehensive conflict weight is determined based on the conflict detection result under the constraint of the corrected main chain;According to the comprehensive conflict weight, the conflict level and spatial position of each planning layer are marked. According to the scheme of the application, multiple types of planning conflicts can be detected according to the association relationship between multi-source data, and the logical structure of the planning rules is associated to realize accurate marking and visual analysis.

The application discloses a method for stably identifying acupoints on the back of a human body and an intelligent terminal, and relates to the technical field of acupoint identification. The method comprises the following steps: acquiring an image of the back of a user and acquiring positioning coordinates of a reference acupoint, calculating the proportion of the human body according to the reference acupoint; dividing the reference acupoint into a plurality of groups, determining a reference straight line according to a representative point in each group, and acquiring a spinal alignment identification line; calculating the specific positions of other acupoints according to the proportion of the human body and the spinal alignment identification line. The application acquires an image of the back of a user, labels the coordinates of a reference acupoint, calculates the representative points of a plurality of groups divided by the reference acupoint to determine the reference straight line of the plurality of groups, and then generates a spinal alignment identification line, thereby determining the positions of other acupoints in combination with the proportion of the human body and the spinal alignment identification line. The application comprehensively considers the spinal alignment and the proportion of the human body, realizes accurate labeling of acupoints and improves the identification efficiency, and also ensures the accuracy of the positioning result and the physiological reasonableness.

A building facade and top surface labeling method based on SAR imaging three-dimensional rendering inverse inversion automatically determines projection ranges of a roof and a facade of a building in an SAR image according to a building target three-dimensional point cloud model, is suitable for a building structure with a complex real scene, and is high in practicability. According to the method, automatic processing can be achieved while the destacking problem in the SAR image labeling process is effectively solved, the labeling precision and the fine granularity are correspondingly improved while the labeling efficiency is improved, the method can be popularized to labeling of other target projection ranges of the SAR image, and support is provided for construction of a large-scale high-quality data set of the SAR image.

This utility model discloses a drawing board for engineering cost estimation, relating to the technical field of engineering cost estimation tools. It includes a first drawing board body, a rotating shaft rotatably connected to one side of the first drawing board body, and a second drawing board body fixedly connected to the other side of the rotating shaft. Assembly blocks are symmetrically fixedly connected to one side of the first drawing board body, and fixing components are installed inside the assembly blocks. L-shaped plates are symmetrically slidably connected to one side of both the first and second drawing board bodies, and clamping components are installed on one side of the L-shaped plates. Merging components are installed on both sides of the first drawing board body. This utility model, using the above structure, folds the originally large board surface into a compact form, easily placing it under a desk, in a closet gap, or in a corner of a vehicle, greatly freeing up office and living space. Moreover, the folded drawing board significantly reduces the difficulty of carrying it; whether going out to survey sites or going to construction sites to coordinate drawings, it can be easily packed into a backpack or handbag, meeting the frequent movement needs of engineering cost estimators.

This application relates to a measurement system, measurement method, electronic device, and non-transitory machine-readable storage medium for expressing the morphological development characteristics of the orbital bone wall. This measurement system can automatically identify the orbital contour, superior rectus muscle contour, inferior rectus muscle contour, medial rectus muscle contour, and lateral rectus muscle contour, and output a ratio curve expressing the morphological development characteristics of the orbital bone wall, providing a more valuable reference for the diagnosis of congenital orbital variations, analysis of orbital decompression surgery procedures, and selection of precision.

This application provides a video annotation method, apparatus, electronic device, storage medium, and product, relating to the field of video processing technology. The method includes: acquiring a video to be annotated and an identification result obtained by recognizing the video to be annotated; combining the video to be annotated and the corresponding identification result, using a multimodal large model and prompt words, generating attribute description text for each object in the video to be annotated and scene description text for the video to be annotated, obtaining a description text to be corrected; acquiring a correction result obtained by correcting the description text to be corrected, as the final annotation result of the video to be annotated; acquiring new prompt words obtained by correcting the prompt words, and using the difference between the final annotation result of the video to be annotated and the description text to be corrected, adjusting the parameters of the multimodal large model to obtain a new multimodal large model. This enables fast and accurate video annotation.