Rendering post-capture artificial-reality effects based on artificial-reality state information

1. A method comprising:

retrieving, by a computing device, a video stream that was recorded while a first artificial-reality effect was being displayed on the video stream, wherein each frame of the video stream comprises a real-world scene without the first artificial-reality effect;

retrieving, by the computing device, an artificial-reality state information stream corresponding to the video stream, wherein the artificial-reality state information stream comprises randomness data used for generating one or more non-deterministic features of the first artificial-reality effect while it was being displayed on the video stream;

retrieving, by the computing device, one or more contextual data streams corresponding to the video stream, wherein the first artificial-reality effect displayed on the video stream was rendered based on at least a portion of the one or more contextual data streams;

rendering, by the computing device, a second artificial-reality effect based on at least a portion of the artificial-reality state information stream and a portion of the one or more contextual data streams, wherein the computing device utilizes the retrieved randomness data for rendering one or more non-deterministic features of the second artificial-reality effect; and

displaying, by the computing device, the second artificial-reality effect on the video stream.

2. The method of claim 1 , wherein the one or more contextual data streams comprise a sensor data stream generated by one or more sensors while the video stream is being captured.

3. The method of claim 2 , wherein the one or more sensors comprises one or more of:

an inertial measurement unit (IMU);

an accelerometer;

a device orientation sensor;

a motion sensor;

a velocity sensor;

a device position sensor;

a microphone;

a light sensor;

a touch sensor;

a stylus sensor;

a depth sensor;

a temperature sensor;

a GPS sensor; or

a user input sensor.

4. The method of claim 1 , wherein the one or more contextual data streams comprise a computed data stream generated by an object tracking algorithm performed on content of the video stream.

5. The method of claim 4 , wherein the computed data comprises one or more of:

face recognition data;

face tracking points;

person segmentation data;

object recognition data;

object tracking points;

object segmentation data;

body tracking points;

world tracking points;

a depth;

a point in a three-dimensional space;

a line in a three-dimensional space;

a surface in a three-dimensional space; or

a point cloud.

6. The method of claim 1 , wherein the second artificial-reality effect is determined based on an input of a user associated with the computing device.

7. The method of claim 6 , wherein determining the second artificial-reality effect based on the input of the user comprises:

presenting choices for the second artificial-reality effect to the user;

receiving an indication of a user choice from one or more input sensors associated with the computing device; and

determining the second artificial-reality effect based on the indication of the user choice.

8. The method of claim 7 , wherein the choices for the second artificial-reality effect comprise no artificial-reality effect, the first artificial-reality effect, or one or more artificial-reality effects different from the first artificial-reality effect.

9. The method of claim 1 , wherein the one or more non-deterministic features comprise:

a size of a rain drop;

a path of a rain drop;

a timing of a rain drop;

a size of a snowflake;

a path of a snowflake;

a timing of a snowflake;

a direction of a flying arrow;

a trajectory of a flying arrow;

a timing of a flying arrow;

a size of a bubble;

a moving path of a bubble; or

a moving speed of a bubble.

10. The method of claim 1 , wherein, if the second artificial-reality effect is the first artificial-reality effect, the computing device renders the second artificial-reality effect identical to the first artificial-reality effect based on at least a portion of the artificial-reality state information stream and a portion of the one or more contextual data streams, wherein one or more non-deterministic features of the second artificial-reality effect are generated based at least on the randomness data in the artificial-reality state information stream.

11. The method of claim 1 , further comprising:

receiving, by the computing device, an indication, from one or more input sensors associated with the computing device, that a user associated with the computing device wants to switch to a third artificial-reality effect in the middle of replaying the video stream;

stopping, by the computing device, rendering the second artificial-reality effect on the video stream;

rendering, by the computing device, the third artificial-reality effect based on at least a portion of the artificial-reality state information stream and a portion of the one or more contextual data streams; and

displaying, by the computing device, the third artificial-reality effect on the video stream.

12. The method of claim 11 , wherein the computing device utilizes the randomness data in the artificial-reality state information stream for one or more non-deterministic features of the third artificial-reality effect.

13. The method of claim 1 , wherein the first artificial-reality effect comprises one or more of:

a virtual object;

a three-dimensional effect;

an interaction effect;

a displaying effect;

a sound effect;

a lighting effect; or

a tag.

14. One or more computer-readable non-transitory storage media embodying software that is operable when executed to:

retrieve a video stream that was recorded while a first artificial-reality effect was being displayed on the video stream, wherein each frame of the video stream comprises a real-world scene without the first artificial-reality effect;

retrieve an artificial-reality state information stream corresponding to the video stream, wherein the artificial-reality state information stream comprises randomness data used for generating one or more non-deterministic features of the first artificial-reality effect while it was being displayed on the video stream;

retrieve one or more contextual data streams corresponding to the video stream, wherein the first artificial-reality effect displayed on the video stream was rendered based on at least a portion of the one or more contextual data streams;

render a second artificial-reality effect based on at least a portion of the artificial-reality state information stream and a portion of the one or more contextual data streams, wherein the computing device utilizes the retrieved randomness data for rendering one or more non-deterministic features of the second artificial-reality effect; and

display the second artificial-reality effect on the video stream.

15. The media of claim 14 , wherein the one or more contextual data streams comprise a sensor data stream generated by one or more sensors while the video stream is being captured.

16. The media of claim 15 , wherein the one or more sensors comprises one or more of:

an inertial measurement unit (IMU);

an accelerometer;

a device orientation sensor;

a motion sensor;

a velocity sensor;

a device position sensor;

a microphone;

a light sensor;

a touch sensor;

a stylus sensor;

a depth sensor;

a temperature sensor;

a GPS sensor; or

a user input sensor.

17. The media of claim 14 , wherein the one or more contextual data streams comprise a computed data stream generated by an object tracking algorithm performed on content of the video stream.

18. The media of claim 17 , wherein the computed data comprises one or more of:

face recognition data;

face tracking points;

person segmentation data;

object recognition data;

object tracking points;

object segmentation data;

body tracking points;

world tracking points;

a depth;

a point in a three-dimensional space;

a line in a three-dimensional space;

a surface in a three-dimensional space; or

a point cloud.

19. A system comprising: one or more processors; and a non-transitory memory coupled to the processors comprising instructions executable by the processors, the processors operable when executing the instructions to:

retrieve a video stream that was recorded while a first artificial-reality effect was being displayed on the video stream, wherein each frame of the video stream comprises a real-world scene without the first artificial-reality effect;

retrieve an artificial-reality state information stream corresponding to the video stream, wherein the artificial-reality state information stream comprises randomness data used for generating one or more non-deterministic features of the first artificial-reality effect while it was being displayed on the video stream;

retrieve one or more contextual data streams corresponding to the video stream, wherein the first artificial-reality effect displayed on the video stream was rendered based on at least a portion of the one or more contextual data streams;

render a second artificial-reality effect based on at least a portion of the artificial-reality state information stream and a portion of the one or more contextual data streams, wherein the computing device utilizes the retrieved randomness data for rendering one or more non-deterministic features of the second artificial-reality effect; and

display the second artificial-reality effect on the video stream.