Abstract: According to an aspect, a method for neural rerendering includes obtaining a three-dimensional (3D) model representing a scene of a physical space, where the 3D model is constructed from a collection of input images, rendering an image data buffer from the 3D model according to a viewpoint, where the image data buffer represents a reconstructed image from the 3D model, receiving, by a neural rerendering network, the image data buffer, receiving, by the neural rerendering network, an appearance code representing an appearance condition, and transforming, by the neural rerendering network, the image data buffer into a rerendered image with the viewpoint of the image data buffer and the appearance condition specified by the appearance code.

Abstract: A bone foundation guide system for a dental implant surgical site has a bone foundation guide and a dental implant surgical guide with an open surgical space therebetween. The top of the foundation guide body is contoured and has an upper surface to attach a bottom surface of the dental implant surgical guide as well as to define a removal and augmentation level and to guide the removal and/or augmentation of bone segments from the dental surgical site via a separation plane. The separation plane has a central area predetermined to define an implant depth of bores for anchoring implants, and further has an inclined area extending from all side ends of the central area towards the first bridge end and the second bridge end raising above the removal and augmentation level of the jaw bone.

Abstract: Embodiments provide techniques for rendering augmented reality effects on an image of a user's face in real time. The method generally includes receiving an image of a face of a user. A global facial depth map and a luminance map are generated based on the captured image. The captured image is segmented into a plurality of segments. For each segment in the plurality of segments, a displacement energy of the respective segment is minimized using a least square minimization of a linear system for the respective segment. The displacement energy is generally defined by a relationship between a detailed depth map, the global facial depth map and the luminance map. The detailed depth map is generated based on the minimized displacement energy for each segment in the plurality of segments. One or more visual effects are rendered over the captured image using the generated detailed depth map.

Abstract: Embodiments relate to an aligner breakage solution. A method includes obtaining a digital design of a polymeric aligner for a dental arch of a patient. The polymeric aligner is shaped to apply forces to teeth of the dental arch. The method also includes performing an analysis on the digital design of the polymeric aligner using at least one of a) a trained machine learning model, b) a numerical simulation, c) a geometry evaluator or d) a rules engine. The method may also include determining, based on the analysis, whether the digital design of the polymeric aligner includes probable points of damage, wherein for a probable point of damage there is a threshold probability that breakage, deformation, or warpage will occur. The method may also include, responsive to determining that the digital design of the polymeric aligner comprises probable points of damage, performing corrective actions based on the probable points of damage.

Abstract: A method is disclosed for spinal anatomy segmentation. In one example, the method includes combining a fully convolutional network with a residual neural network. The method also includes training the combined fully convolutional network with the residual neural network from end to end. The method also includes receiving at least one medical image of a spinal anatomy. The method also includes applying the fully convolutional network with the residual neural network to at least one medical image and segmenting at least one vertebral body from the at least one medical image of the spinal anatomy.

Abstract: Methods and systems for determining orthodontic treatment based on 3D digital model of subject's teeth. The method includes determining, for a given trajectory segment of the orthodontic treatment of a given tooth in which the given tooth is moved from a start position to an end position, whether the force required to move the given tooth from the start position to the end position causes an induced stress associated with at least one other of the plurality of subject's teeth. In response to the induced stress being outside of a predetermined threshold level, determining a counter force to be applied to the at least one other of the plurality of subject's teeth. The orthodontic treatment can be thus determined as including the force to be applied to the given tooth and the determined counter force to be applied to the at least one other of the plurality if subject's teeth.

Abstract: A learning-based clothing animation method and system for highly efficient virtual try-on simulations is provided. Given a garment, the system preprocess a rich database of physically-based dressed character simulations, for multiple body shapes and animations. Then, using a database, the system trains a learning-based model of cloth drape and wrinkles, as a function of body shape and dynamics. A model according to embodiments separates global garment fit, due to body shape, from local garment wrinkles, due to both pose dynamics and body shape. A recurrent neural network is provided to regress garment wrinkles, and the system achieves highly plausible nonlinear effects, in contrast to the blending artifacts suffered by previous methods.

Abstract: Apparatuses, systems, and techniques to identify a shape or camera pose of a three-dimensional object from a two-dimensional image of the object. In at least one embodiment, objects are identified in an image using one or more neural networks that have been trained on objects of a similar category and a three-dimensional mesh template.

Abstract: A method includes processing an input comprising data from an intraoral image using a trained machine learning model that has been trained to classify regions of dental sites, wherein the trained machine learning model outputs a probability map comprising, for each pixel in the intraoral image, a first probability that the pixel belongs to a first dental class and a second probability that the pixel belongs to a second dental class, wherein the first dental class represents excess material, the excess material comprising material other than teeth or gums. The method further includes determining, based on the probability map, one or more pixels in the intraoral image that are classified as excess material. The method further includes hiding or removing from the intraoral image data for the one or more pixels that are classified as excess material.

Abstract: A method of generating and modifying a 2D image of a 3D scene the method including the steps: processing an image of a 3D scene to generate a set of data points representative of the 3D scene and 3D objects within the scene; retrieving one or more data points from the set of data; transforming the one or more data points according to one or more mathematical conversion functions, including; a function defining a projection trajectory for each data point; a function defining a geometry of a projection surface for each data point; a function defining a projection volume for each data point; a function defining an angle of projection of each data point with respect to a convergence point on a projection surface; a function defining the size to distance ratio of each data point from a projection surface; generating a transformed set of data points; projecting the transformed set of data points representative of a modified 2D image of the 3D scene; and, rendering the projected transformed set of data points into a

Abstract: Methods and apparatus relating to three-dimensional object models are described. In one example, (i) data representing a geometrical description of a three-dimensional object defining object geometry in a geometric space and (ii) at least one object property description describing an object property in an object property space are received. The object property space and the geometric space are intersected to define an object model, wherein an object property is defined at an intersection between a described object property and defined object geometry.

Abstract: Techniques are described for, given two images taken at two viewing locations, generating and presenting an image transition sequence that simulates viewer movement between the viewing locations, such as without using information about relative distance between the viewing locations and other depth information to objects in an environment around the viewing locations. The start and end images for an image transition sequence may be part of panorama images acquired at the viewing locations, and generating of the image transition sequence may include selecting and using one of multiple types of transition sequences to perform rotating (e.g., from a start image's viewing direction to face toward the viewing location for the end image), zooming (e.g., to simulate moving forward), and/or blending of the end image with a modified version of the start image, with the resulting image transition sequence played as a video or other animation.

Abstract: Described herein are systems, methods, and computer readable media for performing data conversion on sensor data to obtain modified sensor data that is formatted/structured appropriately for downstream processes that rely on the sensor data as input. The sensor data can include point cloud data captured by a LiDAR, for example. A grid structure and corresponding grid characteristics can be determined and the sensor data can be converted to grid-based sensor data by associating the grid structure and its characteristics with the sensor data. Generating the grid-based sensor data can include reformatting the point cloud data to superimpose the grid structure and its grid characteristics onto the point cloud data. Various downstream processing that cannot feasibly be performed on the raw sensor data can then be performed efficiently on the modified grid-based sensor data by virtue of the grid structure imbuing the sensor data with spatial proximity information.

Abstract: An information processing apparatus includes a first detection unit configured to detect a first object located on a first direction side of a refracting surface from an image obtained by an imaging apparatus located on the first direction side, a second detection unit configured to detect a second object located on a second direction side of the refracting surface from an image obtained by an imaging apparatus located on the second direction side, an obtaining unit configured to obtain position information indicating at least either one of a positional relationship between the first and second objects and positions of the first and second objects, and an identification unit configured to identify positions of the plurality of imaging apparatuses in a common coordinate system, based on a result of detection performed by the first detection unit, a result of detection performed by the second detection unit, and the position information.

Abstract: A method and a system for manufacturing an orthodontic appliance are provided. The method comprises: receiving a 3D mesh including a plurality of inner vertices representative of an inner surface of the appliance; generating a reference plane positioned relative to the arch form 3D mesh according to a predetermined position; determining, based on the 3D mesh, for each one of the plurality of inner vertices, a respective distance to the reference plane, the respective distance being indicative of a thickness of the appliance after forming; generating, based on the respective distance, a plurality of outer vertices representative of an outer surface of the appliance; generating, based on the plurality of inner vertices and the plurality of outer vertices, an appliance 3D representation of the appliance including data indicative of the thickness of the appliance; causing the manufacturing of the appliance based on the appliance 3D representation.

Abstract: An apparatus comprises a receiver (201) for receiving an image signal comprising a number of three dimensional images representing a scene from different viewpoints, each three dimensional image comprising a mesh and a texture map, the signal further comprising a plurality of residual data texture maps for a first viewpoint being different from the different viewpoints of the number of three dimensional images, a first residual data texture map of the plurality of residual texture maps providing residual data for a texture map for the first viewpoint relative to a texture map resulting from a viewpoint shift of a first reference texture map being a texture map of a first three dimensional image of the number of three dimensional images and a second residual data texture map of the plurality of residual texture maps providing residual data for a texture map for the first viewpoint relative to a texture map resulting from a viewpoint shift of a second reference texture map being a texture map of a second three

Abstract: The present disclosure relates to an image processing apparatus and an image processing method capable of storing auxiliary information in CbCr components in a YCbCr format. The image processing apparatus includes a receiving section that receives depth image data in which a depth image transmitted together with a texture image is stored in a Y component in a YCbCr format and auxiliary information is stored in CbCr components in the YCbCr format, and an auxiliary information utilization section that executes a predetermined image process using the auxiliary information on at least one of the texture image or the depth image. A value that can be taken on by the auxiliary information for each pixel of the texture image has N patterns, and a gradation value out of N×N gradation values into which a combination of the auxiliary information regarding two pixels is converted is stored in the CbCr components.

Abstract: A method for measuring shoe comfort level is provided. A 3D polygonal model of a foot is constructed from an array of images taken from multiple angles of the foot and leg. Each vertex point is loaded with foot anthropometric sensitive characteristics information. A 3D model of a shoe's inner space is constructed with a scanning device having various opening modes. Each of shoe inner space vertex points is complimented by shoe data such as material type, rigidity, structure, vamp and shoelaces. The 3D polygonal foot model and the vertex inner shoe space model are fitted together and the comfort level for the foot is defined with the simulation of movement.

Abstract: Systems and methods are disclosed for secret sharing for secure collaborative graphical design. Graphical secret shares are generated from a three-dimensional graphical design and distributed to one or more contributor devices. Contributor graphical designs modifying graphical secret shares may be received from contributor devices. Various corresponding and related systems, methods, and software are described.

Abstract: An animation system wherein a machine learning model is adopted to generate animated facial actions based on parameters obtained from a live actor. Specifically, the anatomical structure such as a facial muscle topology and a skull surface that are specific to the live actor may be used. A muscle structure of simplified “pseudo” muscles that approximate the actual muscle topology but with reduced degree of freedom is determined to improve computational efficiency.