Machine learning-based player controls in fully controlled camera systems

Abstract

Systems and methods described herein relate to systems and methods for managing player controls within a fully controlled camera system for an interactive environment. The system automatically renders camera views from different orientations, creating cinematic cuts without direct player input. In response to a cut from a first orientation to a second, the system establishes a new control scheme relative to the new view to maintain intuitive gameplay. A transition to the second control scheme is delayed based on a predicted adaptation time generated by a machine-learning model. This model analyzes the current game state context to determine how long a specific player needs to reorient. Additionally, the system utilizes player hints provided via input devices, converting these hints into contextual weights via a weighting model. These weights influence the scoring of potential camera points, allowing the player to guide camera selection without disrupting the cinematic experience.

Dependencies are inferred from the claim text.