Abstract: Techniques are described for training a machine learning (ML) model is learn compression errors for various compression modes of BC7 given an input set of features that depend on per-channel pixel value ranges in a BC7 block.

Abstract: System and method that includes: receiving a 3D textured model of a part that includes inspection data collected for an industrial process associated with the part; tracking a viewing perspective of a user relative to a reference part; generating pose data based on the tracked viewing perspective; rendering an image that includes inspection data from a 3D textured model based on the generated pose data.

Abstract: Provided are an image processing method and an image processing apparatus. The image processing method includes obtaining scan data by scanning an object including a hole, obtaining at least one parameter for recognizing an inner region of the hole from the scan data, and recognizing the inner region of the hole from the scan data based on the obtained at least one parameter.

Abstract: The present invention relates to a system and method of providing a digital asset marketplace. In particular, the present invention relates to facilitating buying and selling of Non-Fungible Tokens (NFTs) by providing NFT owners with a means of showcasing and offering their NFTs for purchase in a preferred location or space, and similarly, providing potential purchasers with the ability to locate, view and purchase or trade NFTs according to their preferences.

Abstract: An example device for decoding point cloud data includes a memory configured to store point cloud data; and one or more processors implemented in circuitry and configured to: determine a number of times to upscale a downscaled encoded representation of a point cloud geometry; decode the downscaled encoded representation of the point cloud geometry; upscale the downscaled representation of the point cloud geometry the number of times to form an upscaled representation of the point cloud geometry; and reproduce a point cloud using the upscaled representation of the point cloud geometry.

Abstract: Described herein is a method to segment meshes into sub-meshes based on triangle properties. The triangles are first classified according to some characteristic using their respective areas. A filtering process may change the classification according to the neighboring triangles. Then connected components are generated, and neighboring connected components are merged following a certain criteria. In some embodiments, connected components that share the most amount of edges are merged. With this technique, sub-meshes can be automatically generated without any previous knowledge of the mesh generation stage.

Abstract: A method, performed by a health management server, for evaluating a health condition by using a skeleton model. In detail, the method comprises the steps of: providing, to a user apparatus, a standard motion image of a trainer performing a standard motion; receiving, from the user apparatus, a user motion image of a user motion performed by a user following the standard motion; calculating a similarity between the standard motion and the user motion by comparing the standard motion image with the user motion image; and evaluating a health condition of the user on the basis of the calculated similarity.

Abstract: Example devices, systems, and techniques are described. An example technique includes determining that resampling is to be applied to a first reference frame for a slice of point cloud data or a frame of the point cloud data. The technique includes applying resampling to the first reference frame to generate a resampled reference frame. The technique includes determining one or more inter prediction candidates based on the resampled reference frame. The technique includes processing the slice of the point cloud data or the frame of the point cloud data based on the one or more inter prediction candidates.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating compressed representations of synthetic images. One of the methods is a method of generating a synthetic image using a generative neural network, and includes: generating, using the generative neural network, a plurality of coefficients that represent the synthetic image after the synthetic image has been encoded using a lossy compression algorithm; and decoding the synthetic image by applying the lossy compression algorithm to the plurality of coefficients.

Abstract: An online system generates images for collections of items using an image generation model. To ensure a generated image accurately reflects a collection of items, the online system determines a type of the collection and selects a template including evaluation questions associated with the determined type. Evaluation questions are curated to determine accuracy of the content of a generated image for the collection. By applying a visual learning model to the questions in the selected template and the generated image, the online system identifies discrepancies between the image and the collection of items from the output of the vision language model. Subsequently, the online system prompts the image generation model to create an updated image for the collection that does not include the identified discrepancies. The online system may repeat the discrepancy identification and image modification until no discrepancies are found in the generated image.

Abstract: Information is received by one or more servers regarding a plurality of components of a virtual environment to be presented for display by a remote client device. A set of remote rendering prioritization values, indicating a degree of prioritization for rendering individual components of the plurality of components by the one or more server computing devices, is determined based on the received information. Based on the set of remote rendering prioritization values, a partial scene of the virtual environment is generated at the one or more servers by rendering a set of first components of the plurality of components. The generated partial scene of the virtual environment is transmitted to the remote client device for compositing with a set of second components of the plurality of components to be rendered by the remote client device.

Abstract: A method of encoding a mesh includes determining a symmetry plane that divides the mesh into a first side and a second side opposite the first side, the mesh comprising a plurality of vertices, the first side comprising a first set of vertices, the second side comprising a second set of vertices; determining, for each vertex in the first set, a predicted vertex on the second side that is symmetric to a corresponding vertex in the first set; determining a transformation function that transforms each predicted vertex on the second side, the transformation function determined in accordance with a loss function; applying the transformation function that transforms each predicted vertex on the second side; determining a displacement vector between each transformed predicted vertex and a corresponding vertex in the second set; and generating a bitstream that comprises each encoded vertex in the first set and each displacement vector.

Abstract: In an embodiment, a neuroimager alignment and coupling evaluation (NACE) system, includes a neuroimaging device; a fiducial affixed to the neuroimaging device; an imaging device; and an augmented reality (AR) module coupled to the imaging device and configured to provide alignment data for the neuroimaging device by tracking facial or cranial landmarks and the fiducial.

Abstract: A single image deblurring method via vertical and horizontal decomposition includes receiving, by an input unit of a deblurring network system, a single image; extracting, by an encoder, features from the received single image; decoding, by a horizontal decoder, the features provided from the encoder using a first convolution kernel to generate a first residual image; decoding, by a vertical decoder, the features provided from the encoder using a second convolution kernel generated by rotating the first convolution kernel to generate a second residual image; and synthesizing, by a synthesis unit, the first residual image and the second residual image into the received single image to generate a deblurred image.

Abstract: A method comprises displaying, on a graphical user interface (GUI), a first data visualization comprising a positional representation of a plurality of reaction sites from at least a portion of a substrate, wherein the positional representation of the first data visualization simultaneously indicates relative data quality of the plurality of reaction sites; and altering the first data visualization displayed on the GUI to a second data visualization, the second data visualization comprising the positional representation of the plurality of reaction sites from the at least the portion of the substrate, wherein the positional representation of the second data visualization simultaneously indicates relative data quality of the plurality of reaction sites from at least the portion of the substrate based on a comparison of the data quality values of the plurality of reaction sites an adjusted quality value threshold.

Abstract: This disclosure provides systems, methods, and devices that support image processing. In a first aspect, a method of image processing includes receiving a plurality of rays determined based on a plurality of image frames representing a scene around a plurality of cameras, each of the rays is associated with respective temporal information; determining, from a set of objects depicted by the plurality of image frames, a first object as moving, and a second object as stationary, relative to the plurality of cameras; determining, using a model, a first feature set indicative of the first object; determining, using the model, a second feature set indicative of the second object; and determining a three-dimensional representation of the scene based on the first feature set and the second feature set. Other aspects and features are also claimed and described.

Abstract: This disclosure provides systems, devices, apparatus, and methods, including computer programs encoded on storage media, for mesh stitching for motion estimation and depth from stereo. A processor may obtain an indication of a mesh associated with a set of motion vectors for a first frame and a second frame in a plurality of frames. The processor may detect a set of regions in the mesh including a set of motion disparity values associated with the set of motion vectors, where each motion disparity value in the set of motion disparity values is less than a coherency threshold. The processor may calculate an adjustment to the mesh based on the region. The processor may output an indication of the adjustment to the mesh.

Abstract: Systems and methods of the present disclosure provide acquiring images corresponding to multiple perspectives of a piece of equipment to be modeled. A region of interest in a first image of the images determined to be less blurry than a blur threshold and a brightness of the first image is above a brightness threshold. Based on these threshold relationships a three-dimensional model of the piece of equipment is generated based at least in part on a subset of the images that include the first image.

Abstract: Embodiments of the present disclosure provide an image processing method and apparatus, an electronic device, and a readable storage medium. The method includes: acquiring a to-be-processed image, where the to-be-processed image includes a target object; acquiring, by adopting a feature model, normal feature information and target feature information of the target object, where the target feature information includes: depth feature information and/or tangent feature information; performing, based on the normal feature information and the target feature information, a lighting rendering on the target object in the to-be-processed image to obtain a lighting-rendered-image; and outputting the lighting-rendered-image. Compared with the prior art, in the present disclosure, a terminal device may perform the lighting rendering on the target object based on richer feature information of the target object, which can enrich a lighting rendering effect.

Abstract: A method and a server for adapting a diffusion model are provided. The method comprising: receiving a given plurality of images of a given object; receiving a respective textual description for the given object; generating, based on the given plurality of images, a training set of data including a plurality of training digital objects; sampling, from the plurality of training digital objects, a given reference training digital object for using during a validation step; in cycles, executing the validation step by feeding a given reference training digital object to the diffusion model; in response to a convergence metric associated with a loss function at a given cycle of executing the validation step being below a predetermined threshold, terminating the adapting the diffusion model.