Identity Access Management in Metaverse

Manasi Chhibber¹*, Sarthak Agarwal¹* and Amit Dubey²

Published: September 22, 2022

View Pdf

Abstract  

Metaverse is a virtual apace where people can enter and interact with other users in a computer-generated surroundings. It is an integrated system of 3-dimensional virtual worlds. These are preferably accessed via a VR headset. Users’ eye movements, their voice commands and feedback controllers are mainly used to navigate around the metaverse. In this way they are able to fully immerse themselves in the virtual world and simulate their presence. With the advent of this new technology, challenges are bound to occur. Most importantly challenges related to security of users need to be dealt with. Every person entering and interacting in the metaverse must be a verified user so that identity frauds do not happen, hence keeping the users safe from malicious attacks. To deal with this issue, we have come up with two features, first, an avatar facial tracking based and second, a chat based. By using NLP and Deep Learning, we were able to 73.77% uniquely identity a person in a WhatsApp chat group with other people. For the facial recognition-based security feature, we have just laid out our theoretical approach. In future, we’re planning to make use of other features as well such as speech pattern recognition, body positioning retrieval, retinal movement tracking, etc. for allowing only authorized users to enter and interact in the metaverse.

Keywords: metaverse; avatar security; Natural Language Processing; Deep Learning