Abstract: In some embodiments, a computer-implemented method includes obtaining a without-eyeglasses face scan of a subject, the without-eyeglasses face scan being a three-dimensional (3D) model of a face of the subject without eyeglasses; obtaining a with-eyeglasses face scan of the subject, the with-eyeglasses face scan being a 3D model of the subject with eyeglasses; and using the without-eyeglasses face scan and the with-eyeglasses face scan to place a 3D eyeglasses model on a face model of the subject. In some embodiments of the computer-implemented method, the 3D eyeglasses model is placed on the face model of the subject using frame placement information generated using the without-eyeglasses face scan and the with-eyeglasses face scan.

Abstract: A processor-implemented method with object pose estimation includes: determining an image feature corresponding to a point cloud of an input image; determining semantic segmentation information, instance mask information, and keypoint information of an object, based on the image feature; and estimating a pose of the object based on the semantic segmentation information, the instance mask information, and the keypoint information.

Abstract: Systems and method for an object from a plurality of objects are disclosed. An image of a scene containing the plurality of objects is obtained, and a segmentation map is generated for the objects in the scene. The shapes of the objects are determined based on the segmentation map. An end effector is adjusted in response to determining the shapes of the objects. The adjusting the end effector includes shaping the end effector according to at least one of the shapes of the objects. The plurality of objects is approached in response to the shaping of the end effector, and one of the plurality of objects is picked with the end effector.

Abstract: Example aspects include techniques for implementing a high-frequency attention network for single image super-resolution. These techniques may include extracting a plurality of features from an original image input into a CNN to generate a feature map, and restoring one or more high-frequency details of the original image via an efficient residual block (ERB) and a high-frequency attention block (HFAB) configured to assign a scaling factor to one or more high-frequency areas. In addition, the techniques may include generating reconstruction input information by performing an element-wise operation on the one or more high-frequency details and cross-connection information from the feature map and performing, by the CNN, an enhancement operation on the reconstruction input information to generate an enhanced image.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media that implement an inpainting framework having computer-implemented machine learning models to generate high-resolution inpainting results. For instance, in one or more embodiments, the disclosed systems generate an inpainted digital image utilizing a deep inpainting neural network from a digital image having a replacement region. The disclosed systems further generate, utilizing a visual guide algorithm, at least one deep visual guide from the inpainted digital image. Using a patch match model and the at least one deep visual guide, the disclosed systems generate a plurality of modified digital images from the digital image by replacing the region of pixels of the digital image with replacement pixels. Additionally, the disclosed systems select, utilizing an inpainting curation model, a modified digital image from the plurality of modified digital images to provide to a client device.

Abstract: An assistance method of safe driving applied in a vehicle-mounted electronic device obtains RGB images of scene in front of a vehicle, processes the RGB images by a trained depth estimation model, obtains depth images and converts the depth images into three-dimensional (3D) point cloud maps, determines 3D regions of interest therein, and obtains position and size information of objects in the 3D regions of interest. When the position information satisfies a first preset condition and/or the size information satisfies a second preset condition, the presence of obstacles in the 3D regions of interest is determined and controls the vehicle to issue an alarm. When the position information does not satisfy the first preset condition and/or the size information does not satisfy the second preset condition, the 3D regions of interest are determined as obstacle-free, and permitting the vehicle to continue driving.

Abstract: Systems, computer-implemented methods, and non-transitory machine-readable storage media are provided for detecting recapture attacks of images. One method comprises extracting one or more features from an image captured by a device; applying the one or more features as input to a trained machine learning model, wherein the trained machine learning model outputs a first score based on the extracted features; obtaining metadata of the image; performing a statistical analysis of the metadata of the image; generating a second score based on the statistical analysis of the metadata of the image; and generating a probability that the image is a recapture of an original image based on the first score and the second score.

Abstract: In a computer-implemented method of augmenting a dataset used in facial expression analysis, a first facial image and a second facial image are added to a training/testing dataset and mapped to two respective points in a continuous dimensional emotion space. The position of a third point in the continuous dimensional emotion space between the first two points is determined. Augmentation is achieved when a labelled facial image is derived from the third point based on its position relative to the first and second facial expression.

Abstract: A machine may be configured to perform image evaluation of images depicting items for sale and to provide recommendations for improving the images depicting the items to increase the sales of the items depicted in the images. For example, the machine accesses a result of a user behavior analysis. The machine receives an image of an item from a user device. The machine performs an image evaluation of the received image based on an analysis of the received image and the result of the user behavior analysis. The performing of the image evaluation may include determining a likelihood of a user engaging in a desired user behavior in relation to the received image. Then, the machine generates, based on the evaluation of the received image, an output that references the received image and indicates the likelihood of a user engaging in the desired behavior.

Abstract: A method for generating a plurality of sets of training image data for training a machine learning model includes: (a) acquiring object image data representing an object image; (b) dividing the object image into T number of partial object images by dividing a region of the object image into T number of partial regions corresponding to respective ones of T number of partial color ranges; (c) generating a plurality of sets of color-modified object image data representing respective ones of a plurality of color-modified object images by performing an adjustment process on the object image data, the adjustment process including a color modification process to modify colors of at least one of the T number of partial object images; and (d) generating the plurality of sets of training image data using one or more sets of background image data and the plurality of sets of color-modified object image data.

Abstract: An information processing device (10) acquires a plurality of ledger sheets having the same layout, compares the contents of each column at the same position each of the acquired plurality ledger sheets having the same layout, discriminates the type of each column according to the comparison result, and stores the information on the type of each column in a storage unit (14). Moreover, the information processing device (10) acquires position information of a processing target ledger sheet, compares information on the type of a column and the content of each column using information on a registered style with respect to the acquired ledger sheet, discriminates the type of each column of the processing target ledger sheet according to the comparison result, and specifies style candidates of the processing target ledger sheet on the basis of the discrimination result.

Abstract: An imaging processing method that acquires first and second overlapping image data sets by performing first and second measurements on an overlapping location at first and second times, wherein the first and second times are different times; determines whether the first and second overlapping image data sets have substantially a same image quality; and generating and outputting, a first weighted overlapping combined image by combining (a) first weighted image data generated by applying a first weight to an overlapping frequency range of the overlapping image data set having a higher image quality and (b) second weighted image data generated by applying a second weight to the overlapping frequency range of the overlapping image data set having a lower image quality, wherein the first weight is larger than the second weight.

Abstract: A computer-implemented method is disclosed for segmenting a binarized document. The method includes extracting connected components from the binarized document and discriminating (for at least one of the connected components) whether it is a text component based on a homogeneity level value. The homogeneity level value is representative of the level of homogeneity within the local region of the connected component. The local region includes the connected component and at least one adjacent connected component. The homogeneity level value is based on at least one value representative of at least one image characteristic parameter determined for the connected component and on at least one value representative of the image characteristic parameter of the at least one adjacent connected component.

Abstract: An imaging processing device includes an image acquisition module configured to acquire an image obtained by photographing the sky with a camera having a known orientation or slant; a time stamp acquisition module configured to acquire a photographing date and time of the image; a photographing position acquisition module configured to acquire position information of a photographing position of the image; a sun position determination module configured to determine a photographed sun position in the image; a reference sun position acquisition module configured to calculate a reference sun position indicating a sun position determined based on the photographing date and time and the position information; and a camera information identification module configured to determine any one of unknown orientation and slant of the camera based on any one of the known orientation and slant of the camera, the photographed sun position, and the reference sun position.

Abstract: Provided is an in-vehicle sensor system capable of maintaining an equivalent level of precision (accuracy) of data to be output before and after replacement of a surrounding sensor.

Abstract: The disclosure herein describes preparing and using a cross-attention model for action recognition using pre-trained encoders and novel class fine-tuning. Training video data is transformed into augmented training video segments, which are used to train an appearance encoder and an action encoder. The appearance encoder is trained to encode video segments based on spatial semantics and the action encoder is trained to encode video segments based on spatio-temporal semantics. A set of hard-mined training episodes are generated using the trained encoders. The cross-attention module is then trained for action-appearance aligned classification using the hard-mined training episodes. Then, support video segments are obtained, wherein each support video segment is associated with video classes. The cross-attention module is fine-tuned using the obtained support video segments and the associated video classes.

Abstract: One embodiment provides a method for clustering documents based upon a structure of each of the documents, including: receiving, at a device utilizing the machine-learning model, at least one document, each including a plurality of characters and having a structure; converting, for each of the at least one document, each of the plurality of characters to one of a plurality of character representations, wherein the converting includes identifying an attribute of a character and selecting a character representation corresponding to the attribute; producing at least one array for each of the one or more documents, wherein the at least one array includes the plurality of characters converted to the character representations; and clustering the at least one document into document clusters having similar structures by grouping the at least one arrays into groups of arrays having similarities, wherein each document cluster include documents corresponding to the arrays within one of the groups of arrays.

Abstract: Provided are an artificial intelligence (AI) system using a machine learning algorithm such as deep learning, and applications thereof. A method, performed by an electronic apparatus, of processing images includes obtaining a plurality of training image sets corresponding to a plurality of types of target objects, wherein training images in the training image sets are labeled with feature points forming a preset structure, generating a first artificial intelligence (AI) model for determining a standard structure based on the labeled feature points, by using the training images in the training image sets, identifying a face in a training image transformed based on the standard structure, and training a second AI model for verifying the first AI model, based on an image regressed from the transformed training image, and the training image before being transformed.

Abstract: A method and system for automatic technical assistance is provided. The technical assistant is first trained using the user profile of the users who will be accessing the technical assistant, and the workflow details. The technical assistant continuously monitors user screen and checks if the user is on the correct screen for the required workflow. Based on user confirmation an RPA bot resolves the error for the user. In an embodiment, the bot completes the workflow for the user. The bot may also provide guidance content to the user to help the user realize the error.

Abstract: In some implementations, a device may detect edges in an image, and may identify, based on the edges, a rectangle that bounds a document in the image. The device may detect lines in the image, and may identify edge candidate lines by discarding one or more of the lines. The device may identify intersection points where lines, included in the edge candidate lines, intersect with one another. The device may identify corner candidate points by discarding one or more points included in the intersection points, and may identify a corner point included in the corner candidate points. The corner point may be a point, included in the corner candidate points, that is closest to one corner of the bounding rectangle. The device may perform perspective correction on the image of the document based on identifying the corner point.