Skip to content

AkshatBhat/Action-Based-Language-Learning-For-Duolingo-Using-VR

Repository files navigation

Duolingo World

Action-Based Language Learning with Immersive VR

Problem Statement

Most language-learning platforms rely on memorization and static exercises, leaving learners unprepared for authentic, spontaneous conversations. When faced with real-world situations — ordering at a café, asking for directions — they freeze. Traditional flashcards cannot replicate the speed, accents, and unpredictability of real interactions, making learners feel voiceless and disconnected.

💡 Proposed Solution

We are developing Duolingo World — an action-based language rehearsal feature powered by immersive VR. Learners will engage in authentic, dynamic scenarios, guided by:

  • Functional Language Deployment (practicing language chunks in real context)
  • Contextual Reinforcement (situational learning tied to realistic tasks)
  • Problem-based learning (navigating out-of-stock items, asking for help)

Duolingo World will place learners in realistic VR scenes (like grocery stores) where they interact with NPCs, ask questions, check out, and solve spontaneous challenges using their target language.

🔑 Key elements include:

  • 🥽 Immersive VR environments (Quest 2 hardware)
  • 🗣️ Voice recognition for conversational practice
  • 🎯 Dynamic quests with authentic language

This approach bridges the gap between memorized vocabulary and true language fluency — helping learners live their language, not just memorize it.

🛠️ Tech Stack

🎮 Game Engine

  • Unity

🌐 Backend

  • Flask (Python) – REST API for voice interaction
  • .NET Framework – Supporting C# scripting and Unity integration

🧠 AI & Language Processing

  • Gemini 2.0-Flash API – For LLM-based conversational responses
  • Google Cloud Speech-to-Text – For voice transcription (STT)
  • Google Cloud Text-to-Speech – For audio output (TTS)

☁️ Cloud Platform

  • Google Cloud Platform (GCP) – Hosting APIs and processing services

💻 Programming Languages

  • C# – For Unity scripting and backend interaction
  • Python – For backend logic and integration

🚀 To Launch

  • Install the newest version of Unity 6000.1.9f1
  • Pull this repo, click Add and add from disk, then navigate within VR_Scene within the repo and click Open
  • Click the new reference to the project in the Hub
  • Press Play in the Unity Editor when the project launches with an Oculus Quest setup and connected

🐞 Challenges

  • Allowing the Quest headset to sleep causes Meta Quest Link to disconnect → Unity and Meta Quest Link need to restart
  • VSCode issues had certain C# scripts referring to older versions of other C# scripts leading to invisible bugs
  • Integrating Google Cloud Speech-to-Text and Text-to-Speech APIs with Unity
  • Cross-platform issues with macOS and WindowsOS

🎯 Future Goals

  • Finish detailing the VR scene
  • Integrate the TTS Unity scene with the VR scene via a helper NPC
  • Procedural grocery lists and item location for more thorough learning

📈 Business Impact

Duolingo World can:

  • Deepen user engagement
  • Increase subscription retention
  • Drive higher lifetime value
  • Enhance Duolingo’s brand as a leader in immersive, functional language education
  • Support responsible tourism by helping travelers communicate respectfully and engage with local cultures, addressing rising frustration in many cities toward unprepared, non-communicative visitors

👥 Team

Kohler UIRP Summer Interns 2025:

Aditi Kumar, Akshat Bhat, Caroline Stoklosinski, Deeya Patel, Hongyu Chen, Samuel Hurh

🎥 YouTube Video Link

Watch the video on YouTube

📄 License

This project is for demonstration and educational purposes.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors