A method, apparatus and a product for determining a performance measurement of predictors. The method comprises obtaining a dataset comprising data instances. Each data instance is associated with a label; obtaining a predictor. The predictor is configured to provide a prediction of a label for a data instance; determining a plurality of data slices that are subsets of the dataset. computing, for each data slice in the plurality of data slices and based on an application of the predictor on each data instance that is mapped to the data slice, a performance measurement that is indicative of a successful label prediction for a data instance comprised by the data slice, whereby obtaining a plurality of performance measurements; based on the plurality of performance measurements, computing a performance measurement of the predictor over the dataset; if the performance measurement of the predictor is below a threshold, performing a mitigating action.

Screen response validation of robot execution for robotic process automation (RPA) is disclosed. Whether text, screen changes, images, and/or other expected visual actions occur in an application executing on a computing system that an RPA robot is interacting with may be recognized. Where the robot has been typing may be determined and the physical position on the screen based on the current resolution of where one or more characters, images, windows, etc. appeared may be provided. The physical position of these elements, or the lack thereof, may allow determination of which field(s) the robot is typing in and what the associated application is for the purpose of validation that the application and computing system are responding as intended. When the expected screen changes do not occur, the robot can stop and throw an exception, go back and attempt the intended interaction again, restart the workflow, or take another suitable action.

An oral-area positioning device is provided in the invention. The oral-area positioning device includes a storage device, a positioning circuit and a calculation circuit. The storage device stores information corresponding to a plurality of oral areas. The positioning circuit obtains a target image from an oral-image extracting device, and obtains a first position estimation result according to the information corresponding to the plurality of oral areas and a first algorithm. The positioning device obtains a second position estimation result at least according to the information corresponding to the plurality of oral areas, a second algorithm and a reference image position of a reference image, wherein the reference image position is one of the oral areas. The calculation circuit generates a third position estimation result according to the first position estimation result and the second position estimation result.

Disclosed are a method for simultaneous localization and mapping, device, system and storage medium. The method comprises: acquiring a local binocular vision image and a local three-dimensional laser point cloud image that can be collected at present; preprocessing the local binocular vision image and the local three-dimensional laser point cloud image, respectively; acquiring a position and an attitude of a local binocular vision map according to the preprocessed local binocular vision image; fusing the position and attitude of the local binocular vision map with the preprocessed local three-dimensional laser point cloud image to obtain a local fusion map; and performing a global consistency optimization on a global map according to the local fusion map to obtain a dense point cloud map and output a current position information and a current attitude information in real time so as to complete the simultaneous localization and mapping (SLAM), wherein the global map is constructed according to a global three-dimensional laser point cloud image.

Methods and systems are presented for extracting categorizable information from an image using a graph that models data within the image. Upon receiving an image, a data extraction system identifies characters in the image. The data extraction system then generates bounding boxes that enclose adjacent characters that are related to each other in the image. The data extraction system also creates connections between the bounding boxes based on locations of the bounding boxes. A graph is generated based on the bounding boxes and the connections such that the graph can accurately represent the data in the image. The graph is provided to a graph neural network that is configured to analyze the graph and produce an output. The data extraction system may categorize the data in the image based on the output.

A rail detecting device includes: a starting-region rail-starting-point redesignating unit that shifts a starting region for detecting rails to set a modified starting region and designates modified starting points of two rails at the lower edge of the modified starting region when portions of the two rails are not detected based on rail segments in an image; a rail-candidate detecting unit that detects multiple lines in the modified starting region as multiple modified rail candidates, the lines being candidates of edges of the two rails; a starting-region-rail assigning unit that assigns modified rail candidates whose lowest ends are closest to the corresponding modified starting points out of the multiple modified rail candidates as tentative modified rail segments; and a starting-region-rail disappearance determining unit that establishes the tentative modified rail segments as modified rail segments when distances between the lowest ends of the tentative modified rail segments.

Methods and systems are presented for extracting categorizable information from an image using a graph that models data within the image. Upon receiving an image, a data extraction system identifies characters in the image. The data extraction system then generates bounding boxes that enclose adjacent characters that are related to each other in the image. The data extraction system also creates connections between the bounding boxes based on locations of the bounding boxes. A graph is generated based on the bounding boxes and the connections such that the graph can accurately represent the data in the image. The graph is provided to a graph neural network that is configured to analyze the graph and produce an output. The data extraction system may categorize the data in the image based on the output.

A sensor calibration system configured to receive a first frame of one or more markers on a repositionable platform at a first location within a space from a sensor. The system is further configured to determine pixel locations in the first frame for a first marker and a second marker from among the one or more markers. The system is further configured to receive distance information that corresponds with a distance between the platform and distance measuring devices. The system is further configured to determine (x,y) coordinates for the first marker and the second marker based on the distance information. The system is further configured to generate a homography based on the (x,y) coordinates and pixel locations of the first marker and the second marker. The homography includes coefficients that translate between pixel locations in the first frame of the sensor and (x,y) coordinates in the global plane.

Methods, devices and storage media for data processing are provided. One of the methods include: obtaining a target image, wherein the target image comprises at least one tubular image; determining a spatial distribution feature and an image feature of each of the at least one tubular image; obtaining, based on a tubular structure recognition model, at least one fusion feature respectively corresponding to the at least one tubular image by fusing the spatial distribution feature and the image feature of each of the at least one tubular image; and recognizing, based on the tubular structure recognition model and the at least one fusion feature respectively corresponding to the at least one tubular image, at least one tubular structure respectively corresponding to the at least one tubular image.

A handwritten text processing system processes a digitized document including handwritten text input to generate an output version of the digitized document that allows users to execute text processing functions on the textual content of the digitized document. Each word of the digitized data is extracted by converting the digitized document into images, binarizing the images, and segmenting the images into binary image patches. Each binary image patch is further processed to identify if the word is machine-generated or if the word is handwritten. The output version is generated by combining underlying images of the pages of the digitized document with words from the pages superimposed in a transparent font at positions that coincide with the positions of the words in the underlying images.