Abstract

Abstract

CapCast is a WebRTC based Video conferencing application with the capabilities of delivering captions based on the ongoing conversation of the participating hosts. Our inspiration on development of such an application comes from the fact that none of the available video conferencing applications support the capture of spoken text.

Audio once captured can be used to provide useful services such as captioning, translation & other context based services. Although WebRTC abstracts the developer from the complex process of setting up a connection, managing network topologies, security - encryption etc., developing and analyzing a WebRTC application provides great exposure to networking concepts such as the ICE protocol , Mesh based Peer-to-peer communication , relay based communication etc.

Since our application constitutes a distributed system, we thought of different points of failure throughout the project and adjusted our architecture based on our needs by making various trade-offs.