The training apparatus (2000) performs a first phase training and a second phase training of a discriminator (10). The discriminator (10) acquires a ground-view image and an aerial-view image, and determines whether the acquired ground-view image matches the acquired aerial-view image. The first phase training is performed using a ground-view image and a first level negative example of aerial-view image. The first level negative example of aerial-view image includes scenery of a different type from scenery in the ground-view image. The second phase training is performed using the ground-view image and a second level negative example of aerial-view image. The second level negative example of aerial-view image includes scenery of a same type as scenery in the ground-view image.

An artificial intelligence apparatus for calibrating an output position of a display panel according to an embodiment includes a camera configured to capture an image displayed by the display panel; and a processor configured to: transmit a signal for outputting a position reference image to the display panel, receive, via the camera, a captured image for the display panel, calculate an output position offset for the display panel in a predetermined unit based on the position reference image and the captured image, determine an output position calibration value for the display panel using the calculated output position offset, and transmit the determined output position calibration value to the display panel.

A system and method for three-dimensional (3D) microgeometry and reflectance modeling is provided. The system receives images comprising a first set of images of a face and a second set of images of the face. The faces in the first set of images and the second set of images are exposed to omni-directional lighting and directional lighting, respectively. The system generates a 3D face mesh based on the received images and executes a set of skin-reflectance modeling operations by using the generated 3D face mesh and the second set of images, to estimate a set of texture maps for the face. Based on the estimated set of texture maps, the system texturizes the generated 3D face mesh. The texturization includes an operation in which texture information, including microgeometry skin details and skin reflectance details, of the estimated set of texture maps is mapped onto the generated 3D face mesh.

A two-stage medical image registration method based on progressive images (PIs) to solve the technical problem of low registration accuracy of traditional image registration methods includes: merging a reference image with a floating image to generate multiple intermediate PIs; registering, by a speeded-up robust features (SURF) algorithm and an affine transformation, the floating image with the intermediate PIs to acquire coarse registration results; registering, by the SURF algorithm and the affine transformation, the reference image with the coarse registration results to acquire fine registration results; and comparing the fine registration results of the intermediate PIs, which are acquired by iteration, and selecting an optimal registration result as a final registration image. The method can achieve multimodal registration for brain imaging with MI, NCC, MSD, and NMI superior to those of the existing registration algorithms. The method effectively improves the registration accuracy through the progressive medical image registration strategy.

A system and method for three-dimensional (3D) microgeometry and reflectance modeling is provided. The system receives images comprising a first set of images of a face and a second set of images of the face. The faces in the first set of images and the second set of images are exposed to omni-directional lighting and directional lighting, respectively. The system generates a 3D face mesh based on the received images and executes a set of skin-reflectance modeling operations by using the generated 3D face mesh and the second set of images, to estimate a set of texture maps for the face. Based on the estimated set of texture maps, the system texturizes the generated 3D face mesh. The texturization includes an operation in which texture information, including microgeometry skin details and skin reflectance details, of the estimated set of texture maps is mapped onto the generated 3D face mesh.

The present technology relates to an image processing device, an image processing method, and a program capable of making it easier to recognize standing objects. Movement transformation of moving a subject position where a subject appears in a target image to be processed is performed, depending on a subject distance from a vanishing point in the target image to the subject position. The present technology can be applied to, for example, the image processing and the like of an image taken by a camera unit onboard a vehicle or other moving body.

A surrounding vehicle display device includes: a surrounding information detection device that obtains information on surroundings of a host vehicle; and a vehicle speed sensor that detects a vehicle speed of the host vehicle. The surrounding vehicle display device includes a controller that uses the information obtained by the surrounding information detection device to generate a virtual image that indicates the surroundings of the host vehicle as being viewed from above the host vehicle and a display displays the virtual image. The controller makes a display region of at least a rear region around the host vehicle on the virtual image wide when the vehicle speed detected by the vehicle speed sensor is higher than a low vehicle speed.

Systems and method directed to an inversion-consistent transfer learning framework for generating portrait stylization using only limited exemplars. In examples, an input image is received and encoded using a variational autoencoder to generate a latent vector. The latent vector may be provided to a generative adversarial network (GAN) generator to generate a stylized image. In examples, the variational autoencoder is trained using a plurality of images while keeping the weights of a pre-trained GAN generator fixed, where the pre-trained GAN generator acts as a decoder for the encoder. In other examples, a multi-path attribute aware generator is trained using a plurality of exemplar images and learning transfer using the pre-trained GAN generator.

Aspects of the present disclosure relate to optical systems with ergonomic presentation of content for use in head-worn computing systems. A method for controlling a head-worn computer when viewing virtual images, including image content, that encourages an ergonomic head position to reduce neck pain, includes determining an angle of the head-worn computer relative to horizontal, determining an angle of a line of sight to the center of the virtual image as presented to a user's eye, determining a deviation between the determined angle of the line of sight and a predetermined ergonomic angle, and shifting the image content of the virtual image vertically as displayed to the user's eye so that a portion of the image content is not viewable, wherein the amount of shifting is in reverse correspondence to the magnitude of the determined deviation.

A multi-camera zoom control method and apparatus, and an electronic system and a storage medium. The method includes: in the process of a first camera collecting an image, if the current set magnification input by a user is in a magnification transition zone, starting a second camera; acquiring a corresponding stereo correction matrix on the basis of calibration parameters of the first camera and the second camera; calculating a translation matrix on the basis of an acquired pixel position corresponding relationship between the same content regions of interest that correspond to a first zoomed image and a second zoomed image and in combination with the current set magnification, and then calculating a smooth transition transformation matrix in combination with the stereo correction matrix; and performing, by applying the smooth transition transformation matrix, affine transformation on an image output by the first camera, to obtain a display image of a device.