Abstract: A device including a processor configured to generate, for each of plural query inputs, point information using factors individually extracted from a plurality of pieces of factor data for a corresponding query input and generate pixel information of a pixel position using the point information of points, the plural query inputs being of the points, in a 3D space, on a view direction from a viewpoint toward a pixel position of a two-dimensional (2D) scene.

Abstract: A method and apparatus for neural rendering based on view augmentation are provided. A method of training a neural scene representation (NSR) model includes: receiving original training images of a target scene, the original training images respectively corresponding to base views of the target scene; generating augmented images of the target scene by warping the original training images, the augmented images respectively corresponding to new views of the target scene; performing background-foreground segmentation on the original training images and the augmented images to generate segmentation masks; and training a neural scene representation (NSR) model to be configured for volume rendering of the target scene by using the original training images, the augmented images, and the segmentation masks.

Abstract: An electronic device, from point information and time information, extracts a plurality of pieces of feature data from a plurality of feature extraction models, obtains spacetime feature data based on interpolation of the pieces of feature data, and generate scene information on the target point at the target time instant from the spacetime feature data and a view direction based on the scene information estimation model.

Abstract: Disclosed are a method and device for representing rendered scenes. A data processing method of training a neural network model includes obtaining spatial information of sampling data, obtaining one or more volume-rendering parameters by inputting the spatial information of the sampling data to the neural network model, obtaining a regularization term based on a distribution of the volume-rendering parameters, performing volume rendering based on the volume-rendering parameters, and training the neural network model to minimize a loss function determined based on the regularization term and based on a difference between a ground truth image and an image that is estimated according to the volume rendering.

Abstract: An image deblurring method and apparatus are provided. The image deblurring method includes generating a primary feature representation on a first blur point in an input image and offset information on similar points of the first blur point by encoding the input image by implementing an encoding model, generating secondary feature representations on the similar points by applying the offset information to the primary feature representation, and generating an output image, based on the secondary feature representations and the offset information, by implementing an implicit function model.

Abstract: A computing device and an operation method thereof are disclosed. The method includes unshuffling first image data to generate input data, generating output data by implementing a neural network (NN) model provided the input data, and generating second image data by shuffling the output data.

Abstract: A method with image processing includes: setting an offset window for an offset pattern of a kernel offset and an offset parameter for an application intensity of the kernel offset; determining an output kernel by applying the kernel offset to an input kernel based on the offset window and the offset parameter; and adjusting contrast of a degraded image using the output kernel.

Abstract: A processor-implemented method with degraded image restoration includes: receiving a degraded training image; training a first teacher network of an image restoration network and a second teacher network of the image restoration network to infer differential images corresponding to the degraded training image, wherein each of the first teacher network and the second teacher network comprises a differentiable activation layer and a performance of the first teacher network is greater than that of the second teacher network; initially setting a student network of the image restoration network based on the second teacher network; and training the student network to infer a differential image corresponding to the degraded training image by iteratively backpropagating, to the student network, a contrastive loss that decreases a first difference between a third output of the student network and a first output of the first teacher network and increases a second difference between the third output and a second output

Abstract: A method and apparatus with image correction is provided. A processor-implemented method includes generating, using a neural network model provided an input image, an illumination map including illumination values dependent on respective color casts by one or more illuminants individually affecting each pixel of the input image, and generating a white-adjusted image by removing at least a portion of the color casts from the input image using the generated illumination map.

Abstract: An image restoration method includes determining an anchor image based on individual images of a burst image set, executing a feature extraction network based on the burst image set while using anchor information of the anchor image, and generating a restored image based on a feature map corresponding to an output of the feature extraction network.

Abstract: A modeling method searches for a sequence matched to a user input using a fluid animation graph generated based on similarities among frames included in sequences included in the fluid animation graph and models a movement corresponding to the user input based on a result of the searching. Provided also is a corresponding apparatus and a method for preprocessing for such modeling.

Abstract: A processor-implemented method includes: estimating, from frame images of consecutive frames acquired from one or more sensors, short-term ego-motion information of the one or more sensors; estimating long-term ego-motion information of the one or more sensors from the frame images; determining attention information from the short-term ego-motion information and the long-term ego-motion information; and determining final long-term ego-motion information of a current frame, of the consecutive frames, based on the long-term ego-motion information and the attention information.

Abstract: A processor-implemented method of implementing an attention mechanism in a neural network includes obtaining key-value coupling data determined based on an operation between new key data determined using a first nonlinear transformation for key data of an attention layer, and value data of the attention layer corresponding to the key data; determining new query data by applying a second nonlinear transformation to query data corresponding to input data of the attention layer; and determining output data of the attention layer based on an operation between the new query data and the key-value coupling data.

Abstract: A method and apparatus for displaying a virtual route estimates a position of a vehicle based on sensing data received asynchronously from sensors, and outputs a three-dimensional (3D) virtual route generated by registering a driving environment model corresponding to a driving environment of the vehicle and the position of the vehicle in map information.

Abstract: A vehicle localization method and a vehicle localization apparatus are disclosed. The vehicle localization method includes estimating an initial position of a vehicle based on data sensed by a position sensor, determining a driving lane of the vehicle based on a front-view image captured from the vehicle, correcting the initial position of the vehicle based on the driving lane, and determining a position of the vehicle in the driving lane based on geometry information of a lane boundary in the front-view image.

Abstract: A three-dimensional (3D) rendering method includes extracting sample points from a 3D scene, acquiring rendering result information for the sample points by rendering the sample points, and generating a rendering result image corresponding to an entire rendering based on the rendering result information for the sample points and feature information of the 3D scene.

Abstract: A localization method and apparatus are provided. The method comprises generating a synthetic image corresponding to a current time period based on a three-dimensional (3D) color map corresponding to an image captured at a time period prior to the current time period, and position information at the current time period time, and determining final position information at the current time period based on the synthetic image corresponding to the current time period and an image captured at the current time period.

Abstract: A modeling method searches for a sequence matched to a user input using a fluid animation graph generated based on similarities among frames included in sequences included in the fluid animation graph and models a movement corresponding to the user input based on a result of the searching. Provided also is a corresponding apparatus and a method for preprocessing for such modeling.

Abstract: A method and apparatus for displaying a virtual route estimates a position of a vehicle based on sensing data received asynchronously from sensors, and outputs a three-dimensional (3D) virtual route generated by registering a driving environment model corresponding to a driving environment of the vehicle and the position of the vehicle in map information.

Abstract: A processor-implemented learning method for an image segmentation includes training first duplicate layers, as duplications of trained first layers of a pre-trained model, so that a second feature extracted from a target image by the trained first duplicate layers is matched to a first feature extracted from a training image by the trained first layers; regularizing the trained first duplicate layers so that a similarity between the first feature and a third feature extracted from the training image by the regularized first duplicate layers meets a threshold; and training second duplicate layers, as duplications of trained second layers of the pre-trained model, to be configured to segment the target image based on the regularized first duplicate layers, the trained second layer being configured to segment the training image.