Systems and methods are disclosed for calculating dynamic ambient occlusion (AO) values for character models to yield high-quality approximations of global illumination effects. The approach utilizes a dual component machine-learning model that factorizes dynamic AO computation into a non-linear component, in which visibility is determined by approximating spheres and their casted shadows, and a linear component that leverages a skinning-like algorithm for efficiency. The parameters of both components are trained in a regression against ground-truth AO values. The resulting model accommodates lighting interactions with external objects and can be generalized without requiring carefully constructed training data.

A non-limiting example game apparatus includes a display device, and a game screen is displayed on the display device. For example, a player character, an enemy character, a background object, etc. are displayed in the game screen. If a throw mode is set, a target cursor is moved on a predetermined plane in a virtual space based on an operation of a player, thereby to designate a target object with which a throwing object is made to be collided. When the target object is designated, a route object is displayed on a line segment that connects a throw starting point and a target point that is the designated position on the target object, and a shadow of the route object is displayed.

Systems and methods are disclosed for calculating dynamic ambient occlusion (AO) values for character models to yield high-quality approximations of global illumination effects. The approach utilizes a dual component machine-learning model that factorizes dynamic AO computation into a non-linear component, in which visibility is determined by approximating spheres and their casted shadows, and a linear component that leverages a skinning-like algorithm for efficiency. The parameters of both components are trained in a regression against ground-truth AO values. The resulting model accommodates lighting interactions with external objects and can be generalized without requiring carefully constructed training data.

A non-limiting example game apparatus includes a display device, and a game screen is displayed on the display device. For example, a player character, an enemy character, a background object, etc. are displayed in the game screen. If a throw mode is set, a target cursor is moved on a predetermined plane in a virtual space based on an operation of a player, thereby to designate a target object with which a throwing object is made to be collided. When the target object is designated, a route object is displayed on a line segment that connects a throw starting point and a target point that is the designated position on the target object, and a shadow of the route object is displayed.

A computer-implemented method, system and computer-readable medium for determining an illumination component for a selected point in a multi-dimensional space. The method comprises identifying a set of probes associated with the selected point, the probes located in the multi-dimensional space; for each selected one of the probes, determining which of a plurality of zones for the selected probe contains the selected point and determining visibility of said determined zone from the selected probe; and deriving an illumination component at the selected point by combining scene irradiance data associated with those of the probes from which the corresponding determined zone is determined to be visible. The illumination component being determined may be the diffuse component of global illumination as applicable to computer graphics rendering.

A computer-implemented method, system and computer-readable medium for determining an illumination component for a selected point in a multi-dimensional space. The method comprises identifying a set of probes associated with the selected point, the probes located in the multi-dimensional space; for each selected one of the probes, determining which of a plurality of zones for the selected probe contains the selected point and determining visibility of said determined zone from the selected probe; and deriving an illumination component at the selected point by combining scene irradiance data associated with those of the probes from which the corresponding determined zone is determined to be visible. The illumination component being determined may be the diffuse component of global illumination as applicable to computer graphics rendering.

Various aspects of an apparatus and a method to generate a shadow of an object embedded in a target image are disclosed herein. The method includes generation of an object mask of the embedded object in the target image. The object mask is categorized into a first category or a second category based on a pre-determined set of rules. Based on the categorization, a shadow of the object in the target image is generated. The generation of the shadow comprises dilating the categorized object mask by a predetermined multiplication factor when the categorized object mask corresponds to the first category. The generation of the shadow further comprises determination of depth information of the categorized object mask at a pre-determined height when the categorized object mask corresponds to the second category.

Various aspects of an apparatus and a method to generate a shadow of an object embedded in a target image are disclosed herein. The method includes generation of an object mask of the embedded object in the target image. The object mask is categorized into a first category or a second category based on a pre-determined set of rules. Based on the categorization, a shadow of the object in the target image is generated. The generation of the shadow comprises dilating the categorized object mask by a predetermined multiplication factor when the categorized object mask corresponds to the first category. The generation of the shadow further comprises determination of depth information of the categorized object mask at a pre-determined height when the categorized object mask corresponds to the second category.

A virtual space, in which at least a player character is placed and a back object is placed on a depth side with respect to the player character, is rendered in a frame buffer with orthographic projection and by performing a depth test. A degree of hiding is determined for each pixel on the basis of a hiding determination method based on SSAO, and the hiding determination method is a method in which, when comparing each of depth values of a plurality of sampling points set around a position, in the virtual space, corresponding to each pixel with a depth of a corresponding position in a depth buffer, a first offset is added to a position of the sampling point, and the comparison is performed. Then, a shadow is further rendered on a pixel corresponding to a portion for which the degree of hiding is high.

Systems and methods are disclosed for calculating dynamic ambient occlusion (AO) values for character models to yield high-quality approximations of global illumination effects. The approach utilizes a dual component machine-learning model that factorizes dynamic AO computation into a non-linear component, in which visibility is determined by approximating spheres and their casted shadows, and a linear component that leverages a skinning-like algorithm for efficiency. The parameters of both components are trained in a regression against ground-truth AO values. The resulting model accommodates lighting interactions with external objects and can be generalized without requiring carefully constructed training data.