Abstract: A visual lens system is described that identifies, automatically and without user intervention, digital tool parameters for achieving a visual appearance of an image region in raster image data. To do so, the visual lens system processes raster image data using a tool region detection network trained to output a mask indicating whether the digital tool is useable to achieve a visual appearance of each pixel in the raster image data. The mask is then processed by a tool parameter estimation network trained to generate a probability distribution indicating an estimation of discrete parameter configurations applicable to the digital tool to achieve the visual appearance. The visual lens system generates an image tool description for the parameter configuration and incorporates the image tool description into an interactive image for the raster image data. The image tool description enables transfer of the digital tool parameter configuration to different image data.

Abstract: The present disclosure describes an image rendering method and apparatus, and a related device. The method includes obtaining, by a terminal device, a base image corresponding to a to-be-rendered picture. The terminal device includes a memory storing instructions and a processor in communication with the memory. The method also includes obtaining, by the terminal device, a picture effect indication array corresponding to the base image, the picture effect indication array comprising a picture effect indication value corresponding to a pixel of the base image; rendering, by the terminal device, the base image matching a screen size of the terminal device to obtain a base texture map; and performing, by the terminal device, rendering on the base texture map according to the picture effect indication array. The present disclosure may effectively reduce the quantity of times of drawing in an image rendering process and improve image rendering efficiency.

Abstract: Various implementations disclosed herein include devices, systems, and methods that refine a first 3D surface representation (e.g., a 3D point cloud or a 3D mesh) using a second 3D surface representation that includes a 3D geometric primitive. In some implementations, a first 3D surface representation of a physical environment is obtained including points at 3D locations determined based on data generated by a first sensor. In some implementations, a second 3D surface representation corresponding to at least a portion of the physical environment is obtained that includes at least one 3D geometric primitive. In some implementations, a determination whether to adjust the 3D locations of at least one point of the points of the first 3D surface representation is made based on the 3D geometric primitive, and the 3D locations of the at least one point is adjusted to align with the geometric primitive based on the determination.

Abstract: A method of rendering image frames on a computing system includes rendering an initial frame having a plurality of pixels and determining a low-resolution shading rate for at least one object in the frame or at least a portion of the frame based on at least one metric of the initial frame. The method further includes identifying a shading block having a plurality of pixels based on the low-resolution shading rate and sampling a texture data of the surface associated with the shading block wherein a texture sampling rate is at least two. The average of the texture samples is then applied to the pixels associated with the shading block. The method further includes shading the pixels in the shading block and displaying an output frame on a display device.

Abstract: The present disclosure provides a vector graphics data processing method, system, medium, and vector graphics processing device. The method includes the following operations: building a vector primitive path intersection data structure (PIDS) based on coordinates of path intersections (PIs); when a new PI is generated, comparing information of the new PI to information of existing PIs corresponding to an X coordinate or Y coordinate of the new PI; and storing the information of the new PI at a corresponding position in the PIDS corresponding to the X coordinate or Y coordinate of the new PI based on a result of the comparing. Only effective PI data are saved, thereby reducing memory footprint and memory bandwidth, and improving vector graphics processing performance.

Abstract: Techniques are described for efficiently reducing the amount of total computation in convolutional neural networks (CNNs) without affecting the output result or classification accuracy. Computation redundancy in CNNs is reduced by exploiting the computing nature of the convolution and subsequent pooling (e.g., sub-sampling) operations. In some implementations, the input features may be divided into a group of precision values and the operation(s) may be cascaded. A maximum may be identified (e.g., by 90% probability) using a small number of bits in the input features, and the full-precision convolution may then be performed on the maximum input. Accordingly, the total number of bits used to perform the convolution is reduced without affecting the output features or the final classification accuracy.

Abstract: A method for conducting a three dimensional (3D) video conference between multiple participants, the method may include receiving second participant metadata and first viewpoint metadata by a first unit that is associated with a first participant, wherein the second participant metadata is indicative of a pose of a second participant and an expression of the second participant, wherein the first viewpoint metadata is indicative of a virtual position from which the first participant requests to view an avatar of the second participant; generating, by the first unit, and based on the second participant metadata and the first viewpoint metadata, a second participant representation information; wherein the second participant representation information comprises a compact 3D model of the second participant and a second participant texture map; and determining, for the first participant and during the 3D video conference, a representation of virtual 3D video conference environment, wherein the determining is based

Abstract: A content generation system, the system comprising an input obtaining unit operable to obtain one or more samples of input text and/or audio relating to a first content, an input analysis unit operable to generate n-grams representing one or more elements of the obtained inputs, a representation generating unit operable to generate a visual representation of one or more of the generated n-grams, and a display generation unit operable to generate second content comprising one or more elements of the visual representation in association with the first content.

Abstract: The present disclosure provides a multi-view image decoding method including: obtaining, from a bitstream, three-dimensional geometry information indicating a three-dimensional space of a multi-view image, view independent component information indicating a view independent component, which is uniformly applied to every view, and view dependent component information indicating a view dependent component, which is differently applied according to views; determining a view dependent component for a texture map of a current view from the view dependent component information; generating the texture map of the current view from a view independent component of the view independent component information and the determined view dependent component for the current view; and reconstructing a current view image according to a three-dimensional space that is constructed according to the texture map of the current view and the three-dimensional geometry information.

Abstract: A search space for performing nearest neighbor searches for encoding point cloud data may be trimmed. Ranges of a space filling curve may be used to identify search space to exclude or reuse, instead of generating nearest neighbor search results for at least some of the points of a point cloud located within some of the ranges of the space filling curve. Additionally, neighboring voxels may be searched to identify any neighboring points missed during the trimmed search based on the ranges of the space filling curve.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer-readable media to enhance texture image delivery and processing at a client device. For example, the disclosed systems can utilize a server-side compression combination that includes, in sequential order, a first compression pass, a decompression pass, and a second compression pass. By applying this compression combination to a texture image at the server-side, the disclosed systems can leverage both GPU-friendly and network-friendly image formats. For example, at a client device, the disclosed system can instruct the client device to execute a combination of decompression-compression passes on a GPU-network-friendly image delivered over a network connection to the client device.

Abstract: A stereoscopic imaging device includes at least a first and a second image recording unit configured to record a first and a second original image of an object from different perspectives, wherein the original images differ at least with regard to one item of image information, an image display unit for imaging displayed images, an image processing unit for further processing the original images, and the image processing unit is configured to supplement at least one of the two original images with at least one item of image information from the other original image to generate a displayed image. In addition, a method for generating at least one displayed image that can be imaged on an image display unit is provided.

Abstract: A system including a rendering engine to render a field texture for a field display and a foveal texture for a steerable foveal display and a compositor including a field compositor to generate frames for the field display from the field texture and a foveal compositor to generate frames for the foveal display from the foveal texture. The system further including a composition manager designed to sequence and select what is presented including one or more of data in the field display and the foveal display.

Abstract: Directional pattern generation techniques are described for digital images as implemented by a directional pattern system. In an implementation, a user input is received to specify a direction with respect to the object. A directional pattern system then fills the object using a directional pattern based on the contours of the object as well as the user-specified direction. To do so, the directional pattern system generates a directional vector field that specifies directions with respect to corresponding locations within the field defined by a mesh. Uniform field embedding is employed to transfer the directional vector field to a grid by superimposing the grid onto the mesh of the directional vector field. The directional pattern system then generates the directional pattern within the object by filling the grid with one or more pattern cells.

Abstract: During tracing of a primary ray in a 3-D space (e.g., a 3-D scene in graphics rendering), a ray is found to intersect a primitive (e.g., a triangle) located in the 3-D space. Secondary ray(s) may be generated for a variety of purposes. For example, occlusion rays may be generated to test occlusion of a point of intersection between the primary ray and primitive is illuminated by any of the light(s). An origin for each secondary ray can be modified from the intersection point based on characteristics of the primitive intersected. For example, an offset from the intersection point can be calculated using barycentric coordinates of the intersection point and interpolation of one or more parameters associated with vertices defining the primitive. These parameters may include a size of the primitive and differences between a geometric normal for the primitive and a respective additional vector supplied with each vertex.

Abstract: Disclosed herein is a transmitting method and a receiving method of point cloud data. The transmitting method may include encoding point cloud data, encapsulating a bitstream that includes the encoded point cloud data into a file, and transmitting the file, the point cloud data include at least geometry data, attribute data, or an occupancy map data, the bitstream is stored in multiple tracks of the file, the file further includes signaling data, and the signaling data include spatial region information of the point cloud data.

Abstract: This disclosure describes a system configured to present primary and secondary, tertiary, etc., virtual reality content to a user. Primary virtual reality content may be displayed to a user, and, responsive to the user turning his view away from the primary virtual reality content, a sensory cue is provided to the user that indicates to the user that his view is no longer directed toward the primary virtual reality content, and secondary, tertiary, etc., virtual reality content may be displayed to the user. Primary virtual reality content may resume when the user returns his view to the primary virtual reality content. Primary virtual reality content may be adjusted based on a user's interaction with the secondary, tertiary, etc., virtual reality content. Secondary, tertiary, etc., virtual reality content may be adjusted based on a user's progression through the primary virtual reality content, or interaction with the primary virtual reality content.

Abstract: The present disclosure relates to a tensor-based optimization method for GPU memory management of deep learning, at least comprising steps of: executing at least one computing operation, which gets tensors as input and generates tensors as output; when one said computing operation is executed, tracking access information of the tensors, and setting up a memory management optimization decision based on the access information, during a first iteration of training, performing memory swapping operations passively between a CPU memory and a GPU memory so as to obtain the access information about the tensors regarding a complete iteration; according to the obtained access information about the tensors regarding the complete iteration, setting up a memory management optimization decision; and in a successive iteration, dynamically adjusting the set optimization decision of memory management according to operational feedbacks.

Abstract: Accesses to a mipmap by a shader in a graphics pipeline are monitored. The mipmap is stored in a memory or cache associated with the shader and the mipmap represents a texture at a hierarchy of levels of detail. A footprint in the mipmap of the texture is marked based on the monitored accesses. The footprint indicates, on a per-tile, per-level-of-detail (LOD) basis, tiles of the mipmap that are expected to be accessed in subsequent shader operations. In some cases, the footprint is defined by a plurality of footprint indicators that indicate whether the tiles of the mipmap are expected to be accessed in subsequent shader operations. In that case, the plurality of footprint indicators are set to a first value to indicate that the tile was not access during the first frame or a second value to indicate that the tile was accessed during the first frame.

Abstract: Described are computer implemented techniques to select a single surface of a computer aided design (CAD) as a surface that shares a boundary with a void space that will be discretized, produce a virtual geometry item that is positioned within or about the void region to define the volume to be discretization and propagate a mesh within the defined volume by discretizing the CAD generated geometry by a discretization factor to find a first valid seed point within the CAD generated geometry that satisfies all virtual geometries taken together.

Abstract: Computer-implemented method for transferring style features from at least one source image to a target image, comprising the steps of generating a result image, based on the source and the target image, wherein one or more spatially-variant features of the result image correspond to one or more spatially variant features of the target image; and wherein a texture of the result image corresponds to a texture of the source image; and outputting the result image, and a corresponding device. According to the invention, the texture corresponds to a summary statistic of spatially variant features of the source image.

Abstract: A three-dimensional data encoding method includes encoding information of a current node included in an N-ary tree structure of three-dimensional points included in three-dimensional data, where N is an integer greater than or equal to 2. In the encoding, first information is encoded, the first information indicating a range for one or more referable neighboring nodes among neighboring nodes spatially neighboring the current node, and the current node is encoded with reference to a neighboring node within the range.

Abstract: In various examples, a VPU and associated components may be optimized to improve VPU performance and throughput. For example, the VPU may include a min/max collector, automatic store predication functionality, a SIMD data path organization that allows for inter-lane sharing, a transposed load/store with stride parameter functionality, a load with permute and zero insertion functionality, hardware, logic, and memory layout functionality to allow for two point and two by two point lookups, and per memory bank load caching capabilities. In addition, decoupled accelerators may be used to offload VPU processing tasks to increase throughput and performance, and a hardware sequencer may be included in a DMA system to reduce programming complexity of the VPU and the DMA system. The DMA and VPU may execute a VPU configuration mode that allows the VPU and DMA to operate without a processing controller for performing dynamic region based data movement operations.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing data characterizing a first region in an environment to generate a prediction characterizing a second region in the environment. One of the methods includes obtaining surfel data comprising a plurality of surfels; obtaining sensor data for a plurality of locations in a first region of the environment; determining, from the surfel data, a plurality of first surfels corresponding to respective locations in the first region of the environment; determining, using the first surfels and the sensor data, a difference between i) a first representation of the first region of the environment corresponding to the first surfels and ii) a second representation of the first region of the environment corresponding to the sensor data; and generating a respective object prediction for one or more locations in a second region of the environment.

Abstract: The present disclosure provides a tessellation data processing method, system, medium and vector graphics processing device. The method includes: constructing a data structure including a content table and information tables in memory; when a vector line generated by tessellation intersects an horizontal/vertical line to obtain a new intersection, reading an address and number of Xnodes or Ynodes of an information table in the content table corresponding to a row/column corresponding to the Y/X coordinate of the intersection; according to the address of the information table and the number of X/Ynodes of the information table, reading corresponding X/Ynodes from the memory; comparing information of the intersection with the X/Ynodes, and updating the X/Ynodes in the information table, or adding an X/Ynode to the information table at a position corresponding to the Y/X coordinates.

Abstract: The disclosure relates to compositions containing allyl isothiocyanate and methods of use thereof for stimulating tearing in a subject to treat dry eye.

Abstract: Techniques are described for efficiently reducing the amount of total computation in convolutional neural networks (CNNs) without affecting the output result or classification accuracy. Computation redundancy in CNNs is reduced by exploiting the computing nature of the convolution and subsequent pooling (e.g., sub-sampling) operations. In some implementations, the input features may be divided into a group of precision values and the operation(s) may be cascaded. A maximum may be identified (e.g., by 90% probability) using a small number of bits in the input features, and the full-precision convolution may then be performed on the maximum input. Accordingly, the total number of bits used to perform the convolution is reduced without affecting the output features or the final classification accuracy.

Abstract: Example embodiments relate to methods and devices for generating (quasi-) periodic interference patterns. One embodiment includes a method for generating an interference pattern using multi-beam interference of electromagnetic radiation. The method includes computing a set of grid points in a complex plane representing a grid with a desired symmetry. The method also includes selecting a radius of a virtual circle. Additionally, the method includes selecting a set of grid points in the complex plane that lies on the virtual circle centered around a virtual center point. Further, the method includes associating an argument of each grid point of the selected set of grid points in the complex plane with a propagation direction of plane waves or quasi plane waves or parallel wave fronts. In addition, the method includes obtaining the interference pattern that is a superposition of the plane waves or quasi plane waves or parallel wave fronts.

Abstract: A neural network-based rendering technique increases temporal stability and image fidelity of low sample count path tracing by optimizing a distribution of samples for rendering each image in a sequence. A sample predictor neural network learns spatio-temporal sampling strategies such as placing more samples in dis-occluded regions and tracking specular highlights. Temporal feedback enables a denoiser neural network to boost the effective input sample count and increases temporal stability. The initial uniform sampling step typically present in adaptive sampling algorithms is not needed. The sample predictor and denoiser operate at interactive rates to achieve significantly improved image quality and temporal stability compared with conventional adaptive sampling techniques.

Abstract: Systems, apparatuses, and methods for implementing light volume rendering techniques are disclosed. A processor is coupled to a memory. A processor renders the geometry of a scene into a geometry buffer. For a given light source in the scene, the processor initiates two shader pipeline passes to determine which pixels in the geometry buffer to light. On the first pass, the processor renders a front-side of a light volume corresponding to the light source. Any pixels of the geometry buffer which are in front of the front-side of the light volume are marked as pixels to be discarded. Then, during the second pass, only those pixels which were not marked to be discarded are sent to the pixel shader. This approach helps to reduce the overhead involved in applying a lighting effect to the scene by reducing the amount of work performed by the pixel shader.

Abstract: Images of various views of objects can be captured. An object mesh structure can be created based at least in part on the object images. The object mesh structure represents the three-dimensional shape of the object and includes a mesh with mesh elements. The mesh elements are assigned views first from a subset of views to texture large contiguous portions of the object from relatively few views. Portions that are not textured from the subset views are textured using the full set of views, such that all mesh elements are assigned views. The views first assigned from the subset of views and the views then assigned from the full plurality of views can be packaged into a texture atlas. These texture atlas views can be packaged with mapping data to map the texture atlas views to their corresponding mesh elements. The texture atlas and the object mesh structure can be sent to a client device to render a representation of the object.

Abstract: According to some embodiments, a system, method and non-transitory computer-readable medium are provided comprising a 3D building modeling module; a memory for storing program instructions; a 3D building modeling processor, coupled to the memory, and in communication with the 3D building modeling module and operative to execute program instructions to: receive a region of interest; receive an image of the region of image from a data source; generate a surface model based on the received image including one or more buildings; generate a digital height model; decompose each building into a set of shapes; apply a correction process to the set of shapes; execute a primitive classification process to each shape; execute a fitting process to each classified shape; select a best fitting model; and generate a 3D model of each building. Numerous other aspects are provided.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for transforming patterns of operations on tensors in a computational graph to reduce the memory burden incurred when reshape operations are performed, in particular when deployed to hardware platforms that have vector instructions or vector memory requiring alignment of operands.

Abstract: A marine sonar display device comprises a display, a memory element, and a processing element. The display displays sonar images. The memory element stores sonar data. The processing element is configured to transmit a transmit electronic signal to a frequency steered sonar element which transmits an array of sonar beams into a body of water, each sonar beam transmitted in a different angular direction, receive a receive electronic signal from the frequency steered sonar element, the receive electronic signal including a plurality of frequency components, calculate an array of sonar data slices, one sonar data slice for each frequency component, generate an array of sonar image slices, one sonar image slice for each sonar data slice, and control the display to visually present the array of sonar image slices in near real time and a historical sequence of at least one sonar image slice.

Abstract: A processor receives a request to access one or more levels of a partially resident texture (PRT) resource. The levels represent a texture at different levels of detail (LOD) and the request includes normalized coordinates indicating a location in the texture. The processor accesses a texture descriptor that includes dimensions of a first level of the levels and one or more offsets between a reference level and one or more second levels that are associated with one or more residency maps that indicate texels that are resident in the PRT resource. The processor translates the normalized coordinates to texel coordinates in the one or more residency maps based on the offset and accesses, in response to the request, the one or more residency maps based on the texel coordinates to determine whether texture data indicated by the normalized coordinates is resident in the PRT resource.

Abstract: A method, apparatus and computer program for generating or updating a texture atlas are provided. The texture atlas is generated or updated by obtaining a first texture to be stored in the texture atlas, dividing the first texture into a plurality of parts, storing the plurality of parts in a respective plurality of spaces in the texture atlas, such that the plurality of parts of the first texture may be separated by parts of other textures in the texture atlas, and storing texture reconstruction information defining how to combine the plurality of parts to render the first texture. By dividing the first texture into the plurality of parts, it is possible to store the texture in the texture atlas even in cases where the atlas does not have an available space of the same dimensions as the undivided texture.

Abstract: Aspects of the disclosure provide methods, apparatuses, and a non-transitory computer-readable medium for point cloud compression and decompression. In some examples, an apparatus for point cloud compression/decompression includes processing circuitry. The processing circuitry determines, for each point in a first point cloud, a corresponding point in a second point cloud. The first point cloud is a processed point cloud based on the second point cloud. The processing circuitry determines, for each point in the first point cloud, a corresponding error vector based on the corresponding point in the second point cloud. The processing circuitry determines, for each point in the first point cloud, a corresponding weight factor based on a position of the respective point in the first point cloud. The processing circuitry determines a quality metric for the first point cloud based on the error vectors and the weight factors.

Abstract: Examples are disclosed that relate to computing devices, head-mounted display devices, and methods for displaying holographic objects using slicing planes or volumes. In one example a computing device causes a display system to display a holographic object associated with a holographic volume, the holographic object occluding an occluded holographic object that is not displayed. Location data of at least a portion of a hand is received from a sensor. The location data of the hand is used to locate a slicing plane or a slicing volume within the holographic volume. Based on the location of the slicing plane or the slicing volume, at least a portion of the occluded holographic object is displayed via the display system.

Abstract: A videoconferencing device at an endpoint determines whether a person is a real person standing in front of a display device or if the person is instead an image rendered by a display device. In the first instance the real person will be included in a video feed for transmission to a remote endpoint. In the second instance, images of the display device on which the person is rendered will not be included in the video feed.

Abstract: A picture rendering method, a storage medium, and an electronic apparatus are provided. The method includes: obtaining one or more target images for a two-dimensional (2D) virtual scene, each of the one or more target images including a background, a character object moving in the background image, and a target object located at a fixed position in the background; determining a first depth value of a first pixel on the target object in the background by using a depth offset indicated in a depth map corresponding to the target object, the depth map recording a depth offset of each first pixel on the target object; obtaining a second depth value of a second pixel on the character object moving to a target display position; and rendering the one or more target images according to the first depth value and the second depth value.

Abstract: Methods and graphics processing systems render items of geometry using a rendering space which is subdivided into a plurality of first regions. Each of the first regions is sub-divided into a plurality of second regions. Each of a plurality of items of geometry is processed by identifying which of the first regions the item of geometry is present within, and for each identified first region determining an indication of the spatial coverage, within the identified first region, of the item of geometry, and using the determined indication of the spatial coverage within the identified first region to determine whether to add the item of geometry to a first control list for the identified first region or to add the item of geometry to one or more second control lists for a respective one or more of the second regions within the identified first region.

Abstract: Systems and methods include a server configured to provide an application to a remote computing device. The application displays a project image including a photo of a paintable structure and a hot spot located at a predefined location within the project image and configured to draw attention by at least one of blinking, pulsing, and enlarging. The application displays a first link and a second link in response to a cursor being controlled to touch or roll over the hot spot. The application displays at least one of a color name and a color number associated with a paint color at the predefined location of the hot spot in response to receiving input selecting the first link, and displays a color summary including a list of paint colors used in the project image in response to receiving input selecting the second link.

Abstract: An integrated circuit including a multiplier-accumulator execution pipeline including a plurality of multiplier-accumulator circuits to process the data, using filter weights, via a plurality of multiply and accumulate operations. The integrated circuit includes first conversion circuitry, coupled the pipeline, having inputs to receive a plurality of sets of data, wherein each set of data includes a plurality of data, Winograd conversion circuitry to convert each set of data to a corresponding Winograd set of data, floating point format conversion circuitry, coupled to the Winograd conversion circuitry, to convert the data of each Winograd set of data to a floating point data format. In operation, the multiplier-accumulator circuits are configured to: perform the plurality of multiply and accumulate operations using the data of the plurality of Winograd sets of data from the first conversion circuitry and the filter weights, and generate output data based on the multiply and accumulate operations.

Abstract: Systems, methods, and computer-readable medium are described for operating servers and mobile devices. A mobile device may send a template request including a geographic location to a server device. Based on the received geographic location, the server device may match the template request with a template tagged with a usage location matching the received geographic location. The server device may provide the mobile device with the template wherein the mobile device may combine the template with an image captured at the geographic location. The mobile device may provide the server device with the modified image for distribution to other mobile devices.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for generating and providing a height map of a three-dimensional digital model based on a normal map of the three-dimensional digital model. For example, the disclosed systems can generate a height map from a normal map by utilizing a Poisson model based on Dirichlet boundary conditions. The disclosed systems can determine the Dirichlet boundary conditions for a virtual boundary of a gradient map corresponding to a given normal map by imposing pure Neumann boundary conditions for the virtual boundary. Based on the Dirichlet boundary conditions, the disclosed systems can utilize a Poisson model to determine height values for pixels of a gradient map corresponding to pixels of a given normal map.

Abstract: A method of playing back an image in a display device obtains a stereo image corresponding to both eyes of a user including a left eye and a right eye, obtains positions of the eyes determines warping transform information corresponding to the positions of the eyes based on a database (DB) provided in advance based on a curvature of an display plane of the display device, warps the stereo image by applying the warping transform information to the stereo image, determines pixel values of a display panel included in the display device from the warped stereo image by performing light field rendering based on the positions of the eyes, and reproduces the determined pixel values on the display panel.

Abstract: A three-dimensional (3D) content refactoring system may maintain the external look and behavior of a donor model, while changing the underlying data to conform the donor model to content standards of a modeling application. The 3D content refactoring system may make the donor model adhere to the content standards by mapping the donor model to a target model that is already formatted for the modeling application.

Abstract: Systems and methods are provided for generating an image of a posed human figure or other subject using a neural network that is trained to translate a set of points to realistic images by reconstructing projected surfaces directly in the pixel space or image space. Input to the image generation process may include parameterized control features, such as body shape parameters, pose parameters and/or a virtual camera position. These input parameters may be applied to a three-dimensional model that is used to generate the set of points, such as a sparsely populated image of color and depth information at vertices of the three-dimensional model, before additional image generation occurs directly in the image space. The visual appearance or identity of the synthesized human in successive output images may remain consistent, such that the output is both controllable and predictable.

Abstract: An information processing device includes a displaced position calculation processing unit, a spatial value acquisition processing unit, and a spatial value display processing unit. The displaced position calculation processing unit is configured to determine displaced position data on a position displaced from an optional point on surface shape data on a surface of a measurement target in a given direction by a given distance. The spatial value acquisition processing unit is configured to acquire, based on spatial region values represented by volume data formed in a space containing a shape of the measurement target and the displaced position data, a spatial region value for a position stored in the displaced position data. The spatial value display processing unit is configured to dispose and display a display material corresponding to the spatial region value on the surface shape data.

Abstract: An image processing method is disclosed. The image processing method may include inputting a first image and a third image to a pre-trained style transfer network model, the third image being a composited image formed by the first image and a second image; extracting content features of the third image and style features of the second image, normalizing the content features of the third image based on the style features of the second image to obtain target image features, and generating a target image based on the target image features and outputting the target image by using the pre-trained style transfer network model.