Methods and systems for applying machine learning to generate gameplay audio
AssigneeActivision Publishing, Inc.
InventorMichael Sanders
Life of this patent
Estimating the term…
Abstract
A system and method generate audio reactions for a virtual game character that closely resembles reactions of a human player in a multiplayer video game environment. The method includes capturing reactions of human players and correlating the captured reactions with one or more coinciding game events that have triggered those reactions. The correlated information is used to train a neural network that subsequently aids in generating audio reactions for non-player characters during a similar game event.
Claims20
A video game content publication system for publishing data acquired and/or generated by one or more video game servers, wherein each of the one or more video game servers is configured to host and manage at least one video game and wherein each of the one or more video game servers is in data communication with remotely located client devices configured to execute an instance of the at least one video game, the system comprising:
- one or more first programmatic modules, implemented in the one or more of the video game servers, configured to acquire voice data from a first application executing at least partially in at least one of the remotely located client devices;
- one or more second programmatic modules, implemented in the one or more of the video game servers, configured to acquire at least one first gameplay event matching a first criteria from the instance of the at least one video game executing in the at least one of the remotely located client devices, wherein the instance of the at least one video game is programmatically separate from, and executes independently from, the first application;
- one or more third programmatic modules, implemented in the one or more of the video game servers, configured to correlate a portion of the voice data with the at least one gameplay event;
- a neural network configured to receive and process the at least one gameplay event and voice data correlated with the at least one gameplay event and generate a first audio output associated with the first criteria;
- and
- one or more fourth programmatic modules, implemented on one or more of the plurality of distributed video game servers, configured to provide the first audio output in response to a second gameplay event where the second gameplay matches the first criteria.
A method for video game content publication for publishing data acquired and/or generated by one or more video game servers, wherein each of the one or more video game servers is configured to host and manage at least one video game and wherein each of the one or more video game servers is in data communication with remotely located client devices configured to execute an instance of the at least one video game, the method comprising:
- acquiring voice data from a first application executing at least partially in at least one of the remotely located client devices;
- acquiring at least one first gameplay event matching a first criteria from the instance of the at least one video game executing in the at least one of the remotely located client devices, wherein the instance of the at least one video game is programmatically separate from, and executes independently from, the first application;
- correlating a portion of the voice data with the at least one gameplay event;
- using a neural network to receive and process the at least one gameplay event and voice data correlated with the at least one gameplay event, to generate a first audio output associated with the first criteria;
- and
- providing the first audio output in response to a second gameplay event where the second gameplay matches the first criteria.
Dependencies are inferred from the claim text.