Game event recognition
A game-agnostic event detector can be used to automatically identify game events. Game-specific configuration data can be used to specify types of pre-processing to be performed on media for a game session, as well as types of detectors to be used to detect events for the game. Event data for detected events can be written to an event log in a form that is both human- and process-readable. The event data can be used for various purposes, such as to generate highlight videos or provide player performance feedback.
1. A computer-implemented method, comprising:
receiving video data using an electronic device;
performing, using the electronic device and based at least in part on a configuration file, pre-processing of a frame of the video data to determine an event region;
analyzing, using the electronic device and using one or more event detectors, the video data for the event region;
determining, using the electronic device, an event, based at least in part on a change displayed in the event region, with at least a minimum level of confidence; and
providing, using the electronic device, event data corresponding to the event.
2. The computer-implemented method of claim 1 , further comprising:
determining, using the electronic device, the one or more event detectors using the configuration file, the event detectors performing at least one of image recognition, pattern matching, icon recognition, optical character recognition, audio recognition, or input analysis.
3. The computer-implemented method of claim 1 , further comprising:
determining, using the electronic device, one or more pre-processing algorithms to use for the pre-processing using the configuration file, the pre-processing algorithms performing at least one of region selection, upscaling, downscaling, filtering, stretching, warping, perspective correction, noise removal, color space transform, color isolation, or value thresholding.
4. The computer-implemented method of claim 1 , further comprising:
receiving, using the electronic device from the one or more event detectors, one or more event cues; and
utilizing, using the electronic device, at least one cue-to-event translation algorithm to determine the event based, at least in part, upon the one or more event cues.
5. The computer-implemented method of claim 1 , further comprising:
receiving, using the electronic device, the video data in the form of a file or stream.
6. The computer-implemented method of claim 1 , wherein the event data is written to an event log and associated with a representative video segment found in the video data.
7. The computer-implemented method of claim 6 , further comprising:
analyzing, using the electronic device, the event data in the event log for purposes of identifying events of one or more event types.
8. The computer-implemented method of claim 7 , further comprising:
generating, using the electronic device, related content corresponding to the event data in the event log, the related content including at least one of a highlight video, a montage, coaching, skill determination, or style analysis.
9. The computer-implemented method of claim 1 , further comprising:
analyzing, using the electronic device, additional data for the session, the additional data including at least one of video data associated with a user, audio data associated with the user, biometric data for the user, or user input.
10. A system comprising:
one or more processors; and
memory including instructions that, when executed by the one or more processors, cause the system to:
receive media corresponding to a user session;
determine at least one event region in the media, based at least in part on a configuration file indicating one or more types of pre-processing to be performed on at least a portion of the media;
analyze video data for the at least one event region using a detection engine and one or more event detectors;
determine an event, based at least in part on a change displayed in the at least one event region; and
provide event data corresponding to the event.
11. The system of claim 10 , wherein the instructions when executed further cause the system to:
determine the one or more event detectors using the configuration file, the event detectors performing at least one of image recognition, pattern matching, icon recognition, optical character recognition, audio recognition, or analysis of input from a user in the user session.
12. The system of claim 10 , wherein the instructions when executed further cause the system to:
determine one or more pre-processing algorithms to use for pre-processing of the media using the configuration file, the pre-processing algorithms performing at least one of event region selection, upscaling, downscaling, filtering, stretching, warping, perspective correction, noise removal, color space transform, color isolation, or value thresholding.
13. The system of claim 10 , wherein the instructions when executed further cause the system to:
receive, from the one or more event detectors, one or more event cues; and
utilize at least one cue-to-event translation algorithm to determine the event based, at least in part, upon the one or more event cues.
14. The system of claim 10 , wherein the instructions when executed further cause the system to:
receive the media in the form of a file or stream during, or after, the user session.
15. The system of claim 10 , wherein the instructions when executed further cause the system to:
write the event data to an event log for the user session, the event data being both human and process readable, the event log being accessible at least for identifying events of one or more event types, and associated with a user in the user session.
16. A non-transitory machine-readable medium having stored thereon a set of instructions, which if performed by one or more processors, cause the one or more processors to at least:
receive media representative of a user session, the media including video data;
determine at least one event region in the media, based at least in part on a configuration file indicating one or more types of pre-processing to be performed on at least a portion of the media;
analyze the video data for the at least one event region using one or more event detectors;
determine an event, based at least in part on a change displayed in the at least one event region; and
provide event data corresponding to the event.
17. The non-transitory machine-readable medium of claim 16 , wherein the instructions if performed further cause the one or more processors to:
determine the one or more event detectors using the configuration file, the event detectors performing at least one of image recognition, pattern matching, icon recognition, optical character recognition, audio recognition, or input analysis.
18. The non-transitory machine-readable medium of claim 16 , wherein the instructions if performed further cause the one or more processors to:
determine one or more pre-processing algorithms to use for pre-processing of the media using the configuration file, the pre-processing algorithms performing at least one of event region selection, upscaling, downscaling, filtering, stretching, warping, perspective correction, noise removal, color space transform, color isolation, or value thresholding.
19. The non-transitory machine-readable medium of claim 16 , wherein the instructions if performed further cause the one or more processors to:
receive, from the one or more event detectors, one or more event cues; and
utilize at least one cue-to-event translation algorithm to determine the event based, at least in part, upon the one or more event cues.
20. The non-transitory machine-readable medium of claim 16 , wherein the instructions if performed further cause the one or more processors to:
write the event data to an event log for the user session, the event data being both human and process readable, the event log being accessible at least for identifying events of one or more event types, and associated with a user in the user session.