top of page
Mystical Grove: Speak & Seek 

A whimsical VR escape game exploring voice interaction and presence.

Overview

Mystical Grove is an immersive voice-interactive experience set in an enchanted world that listens, responds, and evolves with you.

 

Players find themselves lost in a glowing forest, guided by a mysterious voice that speaks in riddles. To escape, they must solve clues and help the voice recover three lost relics, each discovered through a blend of speech, sound, and interaction.

​

Designed and developed in Unity using the IBM Watson Speech-to-Text API, the project explores how voice recognition and environmental storytelling can deepen emotional presence in VR.

My Role

Voice Interaction Design, Gameplay Development , Environment Implementation

Bhavya, Aakruti & Sarah

Team

Timeline

November 2021 - December 2021

Unity, IBM Watson SDK, XR Interaction Toolkit , C#

Tools

The Goal

We wanted to test how voice interaction could shape player engagement and emotional immersion.
Our main question was:

“Can speech input make players feel more present and connected to a virtual world?”

vr headset 1.png

Gameplay & Functionality

Players interact with the forest by speaking to the magical voice - solving riddles, triggering animations, and discovering hidden objects.

Voice Recognition:

  • Integrated IBM Watson Speech-to-Text to process player commands.

  • Built a coded dictionary to recognize key words and trigger scene events.

  • Voice commands trigger in-game actions, reinforcing immersion.

Interactions:

  • Used Unity’s XR Interaction Toolkit for natural VR gestures.

  • Supported grabbing, hovering, and striking actions.

  • Combined physical actions with voice triggers to reinforce immersion.

Development Process

Prototyping Interaction

  • Experimented with basic voice commands mapped to object events.

  • Adjusted Watson’s confidence threshold to ensure responsiveness.

Environment Design

  • Created a stylized forest world using soft lighting, mist particles, and ambient sound loops.

  • Balanced visual simplicity with spatial storytelling. The forest “responds” to the player.

Screenshot 2025-10-20 at 5.34.30 PM.png
Screenshot 2025-10-20 at 5.34.06 PM.png

Voice Integration Challenges

  • Speech recognition errors (“two” vs “to”, repeated “yes” command).

  • Compatibility issues between Watson SDK and Unity’s XR Interaction Toolkit.

  • Mitigated through input validation and fallback interaction cues.

Challenges

  • Integrating IBM Watson SDK within a VR framework.

  • Achieving accurate voice recognition for repeated or similar words.

  • Maintaining immersion when speech commands failed to trigger correctly.

Screens

Screenshot 2025-10-20 at 5.16.03 PM.png

Video Demonstration

(Please switch on the volume to listen to the audio interaction!)​

Next Steps

Improve voice recognition accuracy through NLP preprocessing.

Add accessibility features like subtitles or visual captions for voice prompts

Explore IBM sentiment analysis for adaptive emotional responses

Reflection

Mystical Grove: Speak & Seek demonstrates how voice interaction can deepen immersion and accessibility in VR
environments. The project highlighted both the creative potential and technical challenges of integrating speech
recognition in real-time gameplay. User feedback emphasized the engaging atmosphere and intuitive controls,
inspiring future iterations focused on inclusivity and narrative depth.

bottom of page