A beautiful, interactive demo website showcasing a modular and data-efficient framework for Vietnamese voice adaptation in text-to-speech systems. This demo page demonstrates zero-shot and few-shot voice adaptation, cross-language adaptation, and code-switching capabilities.
- Introduction Section: Overview of the framework with language-aware Mixture-of-Experts module, style and prosody conditioning
- Vietnamese Voice Adaptation: Zero-shot and few-shot demos for different Vietnamese accents and voices
- Cross-Language Adaptation: Demonstrations of voice adaptation across languages
- Code-Switch Adaptation: Real-world code-switching examples
demo/
├── index.html # Main HTML file
├── styles.css # Styling
├── script.js # Interactive functionality
├── audio/ # Audio files (prompt and synthesized)
│ ├── *_prompt.wav # Reference prompt audios
│ └── *_syn.wav # Synthesized output audios
└── README.md # This file
Simply open index.html in a web browser. All files are self-contained and work offline.
- Push the
demo/folder to a GitHub repository - Go to repository Settings → Pages
- Select the branch and folder containing the demo
- Your site will be available at
https://username.github.io/repository-name/demo/
- Go to Netlify Drop
- Drag and drop the entire
demo/folder - Get an instant public URL
- Install Vercel CLI:
npm i -g vercel - Navigate to the demo folder:
cd demo - Run:
vercel - Follow the prompts to deploy
Upload the entire demo/ folder to any static hosting service:
- AWS S3 + CloudFront
- Google Cloud Storage
- Azure Static Web Apps
- Any web server
- Colors: Edit CSS variables in
styles.css(:rootsection) - Content: Modify demo data in
script.js(demoDataobject) - Layout: Adjust grid layouts in
styles.css
Works on all modern browsers:
- Chrome/Edge (latest)
- Firefox (latest)
- Safari (latest)
- Mobile browsers
- All audio files should be in the
audio/subdirectory - Audio files are loaded relative to the HTML file location
- The website is fully responsive and works on mobile devices