Skydance Silicon Valley filed 3 patents this quarter, all within the Game Engines category.
The applications focus on automated camera control systems for 3D games that use real-time algorithms and machine learning to manage cinematic viewpoints without manual player input. These Game Engines patents describe methods for dynamically adapting camera orientation during gameplay transitions, positioning cameras based on environmental analysis, and switching between pre-placed viewpoints to maintain player immersion while reducing developer workload.
The Game Engines category contains 3 patents that collectively address the challenge of managing camera perspectives in 3D environments without requiring player intervention. One application describes a system that detects when camera cuts occur during cinematic sequences and automatically adjusts the control scheme to match the new orientation, either instantly or by analyzing how quickly players adapt to the change. Another takes a machine learning approach, using telemetry data to train models that determine optimal camera positioning and characteristics like angles, zoom levels, and transitions for each interactive space, eliminating manual camera control entirely. The third patent introduces a real-time scoring algorithm that evaluates multiple pre-placed viewpoints and switches between them dynamically to maintain proper framing, incorporating a buffer mechanism to prevent jarring transitions and the ability to refine decisions based on historical data and player feedback.