Abstract: A method includes determining whether a region to be recognized corresponding to a specified feature exists in an image to be processed. The method may also include determining a main subject region containing the main subject of the image to be processed in accordance with a preset feature parameter of the main subject region and the coordinates and size of the region to be recognized when the region to be recognized exists. The method may further include extracting an image of the main subject region as a foreground image for an extraction process of a foreground target, and taking the extracted image as the main subject of the image to be processed.

Abstract: Systems and methods for test object classification are provided in which the test object is docked with a target object in a plurality of different poses to form voxel maps. The maps are vectorized and fed into a convolutional neural network comprising an input layer, a plurality of individually weighted convolutional layers, and an output scorer. The convolutional layers include initial and final layers. Responsive to vectorized input, the input layer feeds values into the initial convolutional layer. Each respective convolutional layer, other than the final convolutional layer, feeds intermediate values as a function of the weights and input values of the respective layer into another of the convolutional layers. The final convolutional layer feeds values into one or more fully connected layers as a function of the final layer weights and input values. The one or more full connected layers feed values into the scorer which scores each input vector to thereby classify the test object.

Abstract: Systems and methods for eye image set selection, eye image collection, and eye image combination are described. Embodiments of the systems and methods for eye image set selection can include comparing a determined image quality metric with an image quality threshold to identify an eye image passing an image quality threshold, and selecting, from a plurality of eye images, a set of eye images that passes the image quality threshold.

Abstract: Some implementations may include a method for watermarking an identification document, the method including: receiving a digitally watermarked digital facial portrait of the person, the digital facial portrait embedded with at least one digital watermark identifying payload data linking the digital facial portrait to the person portrayed in the digital facial portrait; receiving results of comparison between the digital facial portrait of the person and the person presenting the digitally watermarked digital facial portrait; retrieving the at least one digital watermark in response to receiving comparison results that the person portrayed in the digital facial portrait is the person presenting the digitally watermarked digital facial portrait; and determining that the digital facial portrait is authentic based on the retrieved at least one digital watermark.

Abstract: A system for creating synthetic data for testing an autonomous system, comprising at least one hardware processor adapted to execute a code for: producing a plurality of synthetic training signals, each simulating one of a plurality of signals simultaneously captured from a common training scene by a plurality of sensors, and a plurality of training depth maps each qualifying one of the plurality of synthetic training signals according to the common training scene; training a machine learning model based on the plurality of synthetic training signals and the plurality of training depth maps; using the machine learning model to compute a plurality of computed depth maps based on a plurality of real signals, the plurality of real signals are captured simultaneously from a common physical scene, each of the plurality of real signals are captured by one of the plurality of sensors, each of the plurality of computed depth maps.

Abstract: A method of determining a composite action from a video clip, using a conditional random field (CRF), the method includes determining a plurality of features from the video clip, each of the features having a corresponding temporal segment from the video clip. The method may continue by determining, for each of the temporal segments corresponding to one of the features, an initial estimate of an action unit label from a corresponding unary potential function, the corresponding unary potential function having as ordered input the plurality of features from a current temporal segment and at least one other of the temporal segments. The method may further include determining the composite action by jointly optimizing the initial estimate of the action unit labels.

Abstract: Systems and techniques are disclosed for labeling objects within an image. The objects may be labeled by selecting an option from a plurality of options such that each option is a potential label for the object. An option may have an option score associated with. Additionally, a relation score may be calculated for a first option and a second option corresponding to a second object in an image. The relation score may be based on a frequency, probability, or observance corresponding to the co-occurrence of text associated with the first option and the second option in a text corpus such as the World Wide Web. An option may be selected as a label for an object based on a global score calculated based at least on an option score and relation score associated with the option.

Abstract: In an automated method of processing fingerprint images, identity information is extracted from prints typically classified as having “no identification value” because of sparse or missing minutiae by capturing ridge information. Bezier approximations of ridge curvature are used as Ridge Specific Markers. Control points arising from Bezier curves generate unique polygons that represent the actual curve in the fingerprint. These polygons are generated for latent prints and reference prints and compared to determine whether a match exists.

Abstract: One or more pieces of product information each having a product identifier and specifying image information are stored in a product information storage unit. A product identifier acquiring unit detects, at least two images that are to be analyzed, that there are images of products corresponding to one or more pieces of product specifying image information, using the stored pieces of product specifying image information, and acquires one or more product identifiers corresponding to the images of the products. A personal attribute value acquiring unit acquires one or more personal attribute values from the least two images that are to be analyzed, and a personal attribute value-related information accumulating unit accumulates the one or more product identifiers acquired by the product identifier acquiring unit and personal attribute value-related information, in association with each other, the product identifiers and the personal attribute values being information acquired from a same image.

Abstract: Techniques are disclosed for estimating patient radiation exposure during computerized tomography (CT) scans. More specifically, embodiments of the invention provide efficient approaches for generating a suitable patient model used to make such an estimate, to approaches for estimating patient dose by interpolating the results of multiple simulations, and to approaches for a service provider to host a dose estimation service made available to multiple CT scan providers.

Abstract: Techniques and systems are described to generate a compact video feature representation for sequences of frames in a video. In one example, values of features are extracted from each frame of a plurality of frames of a video using machine learning, e.g., through use of a convolutional neural network. A video feature representation is generated of temporal order dynamics of the video, e.g., through use of a recurrent neural network. For example, a maximum value is maintained of each feature of the plurality of features that has been reached for the plurality of frames in the video. A timestamp is also maintained as indicative of when the maximum value is reached for each feature of the plurality of features. The video feature representation is then output as a basis to determine similarity of the video with at least one other video based on the video feature representation.

Abstract: An image processing apparatus includes an object detection unit configured to detect an object from an image, a tracking unit configured to track the detected object, a trajectory management unit configured to manage a trajectory of the object being tracked, and a specific object detection unit configured to detect a specific object from the image. In a case where the specific object determination unit detects the object being tracked by the object tracking unit to be the specific object, the trajectory management unit manages a trajectory of the object being tracked at a time point before the time point the object being tracked is detected to be the specific object as the trajectory of the specific object.

Abstract: An image quality assessment and restoration system may include a processor and a memory storing instructions to receive an input image, receive a predetermined number of parameter candidates, generate a reconstructed image from the input image for each parameter candidate, sort the reconstructed images by the overall comparative quality between them and determine the best reconstructed image, calculate the overall comparative qualities between the remaining reconstructed images and the best reconstructed image, eliminate any parameter candidates that are suboptimal based on the calculated overall comparative quality, iteratively generate and sort additional reconstructed images and eliminate suboptimal parameter candidates until each of the remaining parameter candidates is converged, and output the converged parameters for use in image restoration. The overall comparative quality may depend upon the local gradient-based structure information and/or the global texture quality information.

Abstract: The system and method deal with three-dimensional (3D) scanning technology which produces object representations which permit rapid, highly-accurate part identification which is not afforded by traditional two-dimensional (2D) camera imaging. The system and method are applicable to any field wherein repair/replacement parts are needed, such as the plumbing, automotive, fastener, marine, window, door, etc. fields.

Abstract: A method of facial recognition has been developed by the application of a statistical method, standard deviations or standard errors versus sample number plots, to differentiate the degree of reproducibilities of various measurements between facial anthropological landmarks in individual ethnic groups. Reproducible measurements between facial anthropological landmarks in a particular ethnic group mean they are common features shared by individuals of that ethnic group. Non-reproducible measurements are unique features of each individual in that ethnic group which may be used for individual facial recognition purposes. Such methodology may be computerized for automatic facial recognition. A large amount of data of each ethnic group is needed for facial recognition. In turn, the development of databases of each ethnic group will result in a large amount of data of human faces.

Abstract: An image segmentation apparatus comprises a processor which receives user input specifying a boundary of a region of interest of an image. The processor determines speed of the user input. A segmentation component maps the speed of the user input to a width of an area associated with the boundary of the region of interest. The segmentation component is configured to extract training data from the image on the basis of the area, and to train a segmentation model using the extracted training data. The segmentation component is configured to process image elements of the area using the trained segmentation model in order to compute an indication whether the image elements are to be assigned to the region of interest or not.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for associating still images and videos. One method includes receiving a plurality of images and a plurality of videos and determining whether the images are related to the videos. The determining includes, for an image and a video, extracting features from the image and extracting features frames of the video, and comparing the features to determine whether the image is related to the video. The method further includes maintaining a data store storing data associating each image with each video determined to be related to the image.

Abstract: Disclosed techniques relate to forming a block sum of picture elements employing a vector dot product instruction to sum packed picture elements and the mask producing a vector of masked horizontal picture element. The block sum is formed from plural horizontal sums via vector single instruction multiple data (SIMD) addition.

Abstract: A method for exploring an installation space of an article includes acquiring a point cloud of an indoor space including a number of point data with three-dimensional scanning. The method also includes acquiring obstacle data and passage data from the point cloud. The method further includes exploring an installation space of the article to determine a space where an obstacle is placed in the obstacle data and a space where a passage is present, as a space where the article is allowed to be installed.

Abstract: A system and method include obtaining and authenticating image files from users, such as insured users, at the request of an entity, such as an insurance provider. The requesting entity may supply an electronic address of the user and a unique identifier. The system may transmit a link to the electronic address. When selected, the link causes an image authentication application to be installed on a user device. The application takes the images securely and separately from a native camera application. Each image authentication application may be customized for each requesting entity. An authentication server may identify the requesting entity that made the request and identify a corresponding image authentication application to be provided to the electronic address. The images from the image authentication application may be authenticated via reverse image search, time, geolocation, and/or other information. The authenticated images and/or related data may be provided to the requesting entity.