B5 Bachelor Spellweaver- Words of Power  

Team

  • Esther Gülpen Garray
  • Felix Anders
  • Adrian Joos
  • Jana Meißner
  • Luca Hobiger

Supervision

Tamara Voigt, Jonas Ehrhardt

Ever dreamed of being a wizard? In Spellweaver you have to cast every spell by yourself. But be careful, gotta get them right to defeat evil monsters.

Our Goal

If you’ve ever used a VR Headset, you might’ve run into the tedious task of inputting text. This experience is quite similar to navigating a keyboard with the remote of your smart-TV. A viable alternative offered by more and more web services is speech-to-text input. And that was our calling.

We then asked ourselves which STT (speech-to-text) options are there, and how do we implement them into Unity? After all, what better way is there to introduce a feature than by gamifying it. And so we set the following goals:

The game should make it easy for the user to test the performance of each API while seeing exactly what went wrong. It should also make it relatively easy for other developer to implement new API’s from other providers.

menu scene

Introduction

Once started, you have to choose one of the implemented STT-API’s.
After that, you’ll join the wondrous world of Spellweaver - Words of Power. There you’ll find a wand and a book full of all kinds of Spells.
Above both the book and the stand with the wand you’ll find two boards. One will show you the performance of your cast, while the other gives you two play options: sandbox or challenge. Both will spawn different kind of enemy types, but only one will give you a time limit and a final score.

Future

Our main goal would be to add more animations and extended game mechanics to have a more immersive game experience. For example, it would be ideal to make it a tower defense game or to add more types of enemies and spells. A more detailed analytics board with more statistical values like response time would be cool.

Ideally, we’d like to add a second factor onto our analysis board that describes the general understanding of the transcription from a human POV. That however would require a lot of detailed planning, as it’s quite difficult to pinpoint intuition in order to create a viable method.

The team

This is our team of 5 IMI-students from different semesters