| CPC G09B 5/065 (2013.01) | 17 Claims |
|
1. A computer-based method of dynamically converting a virtual reality (VR) classroom into a VR breakout session, the method comprising:
receiving educational content from a VR educational session and profiles of a plurality of users in the VR educational session;
identifying a topic of the educational content;
extracting a plurality of real-world use cases relevant to the identified topic from a knowledge corpus based on the identified topic and the profiles;
determining whether at least two users of the plurality of users have at least one overlapping learning preference based on the profiles; and
in response to determining the at least two users have the at least one overlapping learning preference:
dividing the at least two users having the at least one overlapping learning preference into a subgroup;
converting the VR educational session into a smaller VR breakout session containing at least one real-world use case conforming to the at least one overlapping learning preference of the at least two users in the subgroup, wherein converting the VR educational session into the smaller VR breakout session includes:
feeding extracted keywords from the VR educational session into a generative adversarial network (GAN) model as input;
iterating, until a generator of the GAN model is able to fool a discriminator of the GAN model at least a majority of a time, generating, by the GAN model, a series of images based on the extracted keywords; and
stitching the series of images together to create the smaller VR breakout session;
and
displaying the converted VR educational session to the at least two users in the subgroup.
|