Detecting and responding to rendering of interactive video content

1. Non-transitory data storage storing program instructions executable by one or more processors to carry out operations to detect and respond to rendering of video content by a video presentation device, wherein the video content includes (i) a pre-established video segment that does not vary based on user-interaction during the rendering and (ii) a dynamically-defined video segment that varies based on user-interaction during the rendering, the operations comprising:

obtaining a query fingerprint generated in real-time during the rendering as a representation of the video content being rendered, the query fingerprint including a first portion representing the pre-established video segment and a second portion representing the dynamically-defined video segment;

while obtaining the query fingerprint, (a) detecting a match between the first portion of the query fingerprint and a reference fingerprint that represents the pre-established video segment, (b) based on the detecting of the match, identifying the video content being rendered, (c) after identifying the video content being rendered, applying a trained neural network to at least the second portion of the query fingerprint, and (d) detecting, based on the applying of the neural network, that rendering of the identified video content continues; and

responsive to at least the detecting that rendering of the identified video content continues, taking action specific to the identified video content.

2. The non-transitory data storage of claim 1 , wherein the video content comprises video game content.

3. The non-transitory data storage of claim 2 , wherein detecting that rendering of the identified video content continues comprises detecting that the video content being rendered by the video presentation device continues to be video game content.

4. The non-transitory data storage of claim 2 , wherein identifying the video content being rendered comprises determining an identity of a particular video game being rendered, and wherein detecting that rendering of the identified video content continues comprises detecting that the video content being rendered continues to be the particular video game.

5. The non-transitory data storage of claim 2 , wherein taking action specific to the identified video content comprises causing presentation of supplemental content comprising an offer for additional video game content.

6. The non-transitory data storage of claim 1 , wherein the operations further comprise:

detecting an ending of the detected match,

wherein applying the trained neural network is responsive to at least the detecting of the ending of the detected match.

7. The non-transitory data storage of claim 1 , wherein the operations further comprise:

determining, based on the applying of the trained neural network to at least the second portion of the query fingerprint, that rendering of the dynamically-defined segment of the video content has continued for at least a threshold duration,

wherein taking action specific to the identified video content comprises causing a presentation of supplemental content, and

wherein causing the presentation of supplemental content is further responsive to the determining that rendering of the dynamically-defined segment of the video content has continued for at least the threshold duration.

8. The non-transitory data storage of claim 7 ,

wherein the dynamically-defined segment of the video content comprises interactive video game content, and

wherein causing presentation of the supplemental content further responsive to the dynamically-defined segment of the video content having continued for at least the threshold duration comprises causing presentation of video-game help content.

9. The non-transitory data storage of claim 7 , wherein the pre-established video segment is a first pre-established video segment, wherein the video content includes a second pre-established video segment that also does not vary based on user-interaction during the rendering, and wherein the query fingerprint includes a third portion representing the second pre-established video segment, and

wherein determining, based on the applying of the trained neural network to at least the second portion of the query fingerprint, that rendering of the dynamically-defined segment of the video content has continued for at least the threshold duration comprises detecting a threshold long duration from rendering the first pre-established video segment until rendering the second pre-established video segment.

10. The non-transitory data storage of claim 9 , wherein the match is a first match, the operations further comprising:

detecting a second match between the third portion of the query fingerprint and a reference fingerprint that represents the second pre-established video segment and, based on the detecting of the second match, discontinuing the applying of the trained neural network to the query fingerprint.

11. The non-transitory data storage of claim 1 , wherein obtaining the query fingerprint generated in real-time during the rendering as a representation of the video content being rendered comprises receiving from the video presentation device, via a network, transmissions of the query fingerprint generated in real-time by the video presentation device during the rendering of the video content.

12. The non-transitory data storage of claim 1 , wherein the operations further comprise:

establishing a set of reference data, including the reference fingerprint and the trained neural network, based on computerized analysis of various instances of video game play.

13. The non-transitory data storage of claim 12 , wherein the operations further comprise:

automatically searching for and downloading from a public packet-switched network at least some of the various instances of video game play; and

conducting the computerized analysis on the downloaded various instances of video game play.

14. The non-transitory data storage of claim 13 , wherein the various instances of video game play comprise play-through videos.

15. The non-transitory data storage of claim 1 , wherein the neural network comprises a Long Short Term Memory neural network.

16. Non-transitory data storage storing program instructions executable by one or more processors to carry out operations to detect and respond to playing of a video game, wherein the video game is rendered in real-time on a video display unit, and wherein the video game includes (i) cutscene video segments that do not vary based on user-interaction during the rendering and (ii) interactive video segments that vary based on user-interaction during the rendering, the operations comprising:

obtaining a query fingerprint generated in real-time during the rendering as a representation of the video game being played, wherein the query fingerprint includes (i) a first portion representing a first cutscene video segment and (ii) a second portion representing a first interactive video segment;

detecting a match between the first portion of the query fingerprint and a reference fingerprint that represents the first cutscene video segment and, based on the detected match, identifying the video game being rendered;

after identifying the video game being rendered, applying a trained neural network to at least the second portion of the query fingerprint to detect that the video game being rendered continues to be the identified game; and

responsive to at least detecting that the video game being rendered continues to be the identified video game, causing a presentation of supplemental content.

17. The non-transitory data storage of claim 16 , wherein the operations further comprise:

detecting an ending of the detected match,

wherein applying the trained neural network is responsive to at least the detecting of the ending of the detected match.

18. The non-transitory data storage of claim 16 , wherein the operations further comprise:

determining, based on the applying of the trained neural network to at least the second portion of the query fingerprint, that the first interactive video segment has continued for at least a threshold duration,

wherein causing the presentation of supplemental content is further responsive to the determining that the first interactive video segment has continued for at least the threshold duration, and

wherein the supplemental content comprises video-game help content.

19. A video presentation device comprising:

a video input interface through which to receive video content to be rendered by the video presentation device;

a video presentation interface for presenting the received video content, wherein the video content includes (i) a pre-established video segment that does not vary based on user-interaction during the presenting and (ii) a dynamically-defined video segment that varies based on user-interaction during the presenting;

a network communication interface;

a processing unit;

non-transitory data storage; and

program instructions stored in the non-transitory data storage and executable by the processing unit to carry out operations comprising:

generating a query fingerprint in real-time while the video content is being presented, the query fingerprint representing the video content being presented and including a first portion representing the pre-established video segment and a second portion representing the dynamically-defined video segment,

outputting, for transmission through the network communication interface to a computing system, the first portion of the query fingerprint, wherein the computing system detects a match between the first portion of the query fingerprint and a reference fingerprint that represents the pre-established video segment and, based on the detected match, identifies the video content being presented by the video presentation device,

outputting, for transmission through the network communication interface to the computing system, the second portion of the query fingerprint, wherein, after identifying the video content being rendered, the computing system applies a trained neural network to at least the second portion of the query fingerprint, and detects, based on the applying of the neural network, that the video content being presented by the video presentation device continues to be the identified video content, and

receiving from the computing system, through the network communication interface, from the computing system a directive for the video presentation device to present supplemental content based on the computing system having detected that the video content being presented by the video presentation device continues to be the identified video content.

20. The video presentation device of claim 19 , wherein the video content comprises video game content, wherein the pre-established video segment comprises a cutscene video segment, wherein the dynamically-defined video segment comprises an interactive gameplay video segment, and wherein the supplemental content comprises video-game help content.