This page covers 3 granted patents from EA (1), Gwangju Institute Of Science And Technology (1), and Acezone (1).
EA describes a spectator server architecture that streams customized game state data to unlimited viewers across different devices. Gwangju Institute Of Science And Technology's AI system analyzes multiple human observers to predict which screen areas viewers want to watch, automating camera direction for esports broadcasts. Acezone details a gaming headset that combines active noise cancellation with psychoacoustic masking to prevent players from hearing audience communications during live tournaments.
EA received 1 patent addressing the challenge of streaming live gameplay to massive audiences without overloading game servers. The system separates spectator data processing into dedicated server tiers that generate device-specific game state representations, allowing theoretically unlimited viewers to watch simultaneously. Rather than encoding video streams or routing spectator traffic through game clients, the architecture cascades lightweight game state data across multiple spectator servers that can render the action on everything from high-end PCs to mobile devices.
The Gwangju Institute Of Science And Technology received 1 patent for an AI-driven approach to automatic camera control in esports broadcasts. The system learns from crowdsourced human annotators who indicate which screen regions they want to watch during gameplay, then uses masked region data to predict viewport positions in new situations that were never explicitly programmed. By abstracting in-game objects into multi-channel type representations before processing, the approach handles the visual complexity of game screens more effectively than rule-based camera systems.
Acezone received 1 patent for headset technology that prevents players from extracting useful information from crowd noise at live tournaments. The design layers psychoacoustic speech scrambling on top of active noise cancellation to address the problem of attenuated speech remaining intelligible after standard noise reduction. The system also incorporates game-specific keyword training, acoustic leak detection to prevent tampering, and integration with public address audio feeds to enable predictive cancellation of commentator speech.
All data sourced from USPTO patent filings. Google Patents may take several weeks to index recent publications. If a link is unavailable, search for the patent number at USPTO Patent Public Search.