Abstract: A method for extracting an oil storage tank based on a high-spatial-resolution remote sensing image is provided, including: acquiring an oil storage tank sample, and randomly dividing the oil storage tank sample into a training set and a testing set; building an oil storage tank extraction model based on a Res2-Unet model structure, wherein the Res2-Unet is a deep learning network based on a UNet semantic segmentation structure, and a Res2Net convolution block is configured to change a feature interlayer learning to a granular learning and is arranged in a residual mode; and performing a precision verification on the testing set.

Abstract: An image processing device includes a rotation processor and an image processor. The rotation processor receives an input image and generates a temporary image according to the input image. The image processor is coupled to the rotation processor and outputs a processed image according to the temporary image, wherein the image processor has a predetermined image processing width, a width of the input image is larger than the predetermined image processing width, and a width of the temporary image is less than the predetermined image processing width.

Abstract: Described are methods and systems for training a machine learning (ML) model to detect anomalies in images of documents. A first image of a first set of images of documents is obtained. Each first image relates to a region of the document and the first set of images comprises an image of a document containing an anomaly and an image of a document not containing an anomaly. Signal processing algorithms are applied to the first images to generate a signal for each first image and each algorithm, and a discriminative power of each algorithm is evaluated. Based on the discriminative power, a signal processing algorithm is selected and ML model input data is generated using signals generated by applying the algorithm to second digital images. The ML model is trained using the input data to produce output indicating whether an image of a document contains an anomaly.

Abstract: Devices and methods for capturing vehicle undercarriage images are described. In some instances, a mirror assembly may be used to reflect images of portions of a vehicle undercarriage into a field of view of a camera to be captured, e.g., as a vehicle passes over the mirror assembly. Composite images may be reconstructed from the reflected portions of the vehicle undercarriage, and analysis may be performed on those reconstructed, composite images to identify features in the composite vehicle undercarriage images.

Abstract: The present disclosure relates to the field of artificial intelligence technology, and to a method of estimating a pose of a device and a related device. The method of estimating a pose of a device includes identifying a similar key frame that is similar to a current frame collected by the device from a key frame set, based on a vector distance between the similar key frame and the current frame; obtaining data-related information between image frames based on a feature matching relationship between the current frame and the similar key frame; and obtaining the pose of the device based on the data-related information.

Abstract: A system for high definition (HD) image recognition of criminals is disclosed. The system includes a plurality of cameras, an image recognition server, investigator user devices, a computing device, a database, and a network. At least one processor of the image recognition server is configured to receive a plurality of photographs of a first individual, perform image processing of the plurality of photographs to extract a first set of physical features, store feature data regarding the first set of physical features in the database, receive suspect data regarding a suspected individual from a first investigator user device, match the suspect data with the feature data stored in the database, and transmit an alert to the computing device in the prison, wherein the alert activates a mobile application on each investigator user device to display match data identifying the suspected individual as the first individual based on the feature data.

Abstract: A problem of imbalanced big data is solved by decoupling a classifier into a neural network for generation of representation vectors and into a classification model for operating on the representation vectors. The neural network and the classification model act as a mapper classifier. The neural network is trained with an unsupervised algorithm and the classification model is trained with a supervised active learning loop. An acquisition function is used in the supervised active learning loop to speed arrival at an accurate classification performance, improving data efficiency. The accuracy of the hybrid classifier is similar to or exceeds the accuracy of comparative classifiers in all aspects. In some embodiments, big data includes an imbalance of more than 10:1 in image classes. The hybrid classifier reduces labor and improves efficiency needed to arrive at an accurate classification performance, and improves recognition of previously-unrecognized images.

Abstract: The present disclosure includes a method, apparatus, and non-transitory computer-readable medium for adaptive neural image compression by meta-learning. The method may include generating a substitute input image and a substitute target quality control parameter using an original input image and a target quality control parameter, wherein the substitute input image is a modified version of the original input image and the substitute target quality control parameter is a modified version of the target quality control parameter. The method may further include encoding the substitute input image, based on the substitute input image and the substitute target quality control parameter, using an encoding neural network, to generate a compressed representation of the substitute input image.

Abstract: A vehicle monitoring apparatus includes a hardware processor functioning as an image acquisition unit, an abnormality information acquisition unit, a damage determination unit, and a damage reporting unit. The image acquisition unit serves to acquire a captured image including an image representation of an exterior of a vehicle body of a vehicle. The abnormality information acquisition unit serves to acquire vehicle-body abnormality information indicating an abnormality in the vehicle body. The damage determination unit serves to determine presence or absence of damage to the vehicle body on the basis of the captured image and the vehicle-body abnormality information. The damage reporting unit serves to report, to an information processing apparatus, a damage report including the captured image or an image generated on the basis of the captured image. The damage report is reported in accordance with on a result of the determination on the damage.

Abstract: This application provides an image defect detection method. The method includes obtaining a first image and a second image of a flawless image. A third image is obtained from the second image and the first image, and a fourth image is obtained according to the second image and an image to be detected. A fifth image is obtained based on the third image and the second image. A sixth image is obtained based on the third image and the fourth image. A seventh image is obtained from the fifth image and the sixth image. A defect value of the fourth image is obtained according to the third image and the seventh image. A detection result of the fourth image is determined based on the defect value.

Abstract: Systems and methods are disclosed for identifying a diagnostic feature of a digitized pathology image, including receiving one or more digitized images of a pathology specimen, and medical metadata comprising at least one of image metadata, specimen metadata, clinical information, and/or patient information, applying a machine learning model to predict a plurality of relevant diagnostic features based on medical metadata, the machine learning model having been developed using an archive of processed images and prospective patient data, and determining at least one relevant diagnostic feature of the relevant diagnostic features for output to a display.

Abstract: A geometric calibration apparatus detects points from projection regions onto which markers disposed on a phantom are projected, and calculates an output vector representing a probability distribution that gives a probability with which each point is a projection of each marker, by inputting data corresponding to each point to a learning model. The geometric calibration apparatus extracts a predetermined number of samples based on the probability distribution, obtains a candidate projection matrix by transforming correspondences between markers determined based on the samples among the markers and points determined based on the samples among the points, calculates points into which the markers are transformed by the candidate projection matrix, calculates a difference between a set of the transformed points and a set of the detected points, and designates the candidate projection matrix as a projection matrix when the difference is less than or equal to a threshold.

Abstract: Techniques are provided for improving image data quality, such as in functional imaging follow-up studies, using reconstruction, post-processing, and/or deep-learning enhancement approaches in a way that automatically improves analysis fidelity, such as lesion tracking fidelity. The disclosed approaches may be useful in improving the performance of automatic analysis methods as well as in facilitating reviews performed by clinician.

Abstract: An expression recognition method is described that includes acquiring a face image to be recognized, and inputting the face image into N different recognition models arranged in sequence for expression recognition and outputting an actual expression recognition result, the N different recognition models being configured to recognize different target expression types, wherein N is an integer greater than 1.

Abstract: A discrete attribute value dataset is obtained that is associated with a plurality of probe spots each assigned a different probe spot barcode. The dataset comprises spatial projections, each comprising images of a biological sample. Each image includes a corresponding plurality of discrete attribute values for the probe spots. Each such value is associated with a probe spot in the plurality of probes spots based on the probe spot barcodes. The dataset is clustered using the discrete attribute values, or dimension reduction components thereof, for a plurality of loci at each respective probe spot across the images of the projections thereby assigning each probe spot to a cluster in a plurality of clusters. Morphological patterns are identified from the spatial arrangement of the probe spots in the various clusters.

Abstract: A method, apparatus and computer readable storage to implement an automated system for video surveillance in a casino or other controlled environment. Players in the casino can be automatically scanned and analyzed for whether they are under the legal gambling age or not. When an underage gambler is detected, a casino security employee (or other casino personnel) is notified so they can take the appropriate action. Similarly, players who are excluded from the casino can also be automatically detected and would be ejected when detected.

Abstract: An apparatus for detecting mounting behavior of an animal object includes: a memory that stores a program; and a processor that executes the program. The program extracts animal detection information about an animal object detected from the image by inputting the received image into an animal detection model. Also, the program extracts bounding boxes of which a distance between coordinates of central points is smaller than a first set value, bounding boxes of which a difference in rotational angle is smaller than a second set value, and bounding boxes of which a difference between a vector connecting the central points of the extracted bounding boxes and an orientation of each bounding box is smaller than a third set value. If activity information of the animal object is extracted based on an MHI of the image, it is determined that mounting behavior occurs.

Abstract: The present disclosure provides an apparatus for cylindrical convolutional neural network operation, and an apparatus and method for object recognition and viewpoint estimation using the same. The present disclosure uses a cylindrical convolutional network to recognize objects in the image and to determine the viewpoint, which cylindrical convolutional network performs a convolution operation while sliding the input image in the direction of rotation of the cylindrical kernel and extracts a plurality of view-specific feature vectors according to angular unit intervals, to identify and recognize objects, as well as to determine the viewpoint from which the objects were photographed.

Abstract: There is provided a method for generating a learning model, the method including: acquiring an endoscopic image captured by an endoscope and manipulation information regarding a manipulation of an endoscope operator in each stage of operation of the endoscope by the endoscope operator operating the endoscope; and generating a learning model learned so as to output the manipulation information of a next stage in a case where the endoscopic image and the manipulation information are input, based on training data including the acquired endoscopic image and manipulation information, and the manipulation information of the next stage.

Abstract: An adaptive image shading correction method and an adaptive image shading system are provided. The method includes: configuring an image capturing device to obtain a current frame; and configuring a processing unit to: divide the current frame into blocks; select block pairs from the blocks, in which each of the block pairs includes an inner block and an outer block; perform a filtering process for each of the block pairs to determine whether a brightness condition, a saturation condition, a hue similarity condition, and a sharpness similarity condition are met; in response to obtaining filtered block pairs, calculate a sum similarity threshold based on hue statistical data, a saturation difference, and a brightness difference; and use filtered blocks with individual thresholds less than the sum similarity threshold to calculate a shadow compensation value to adjust the current frame.