Open-Source AI Research & Creative Tools Platform
Developer: Hasan Arthur Altuntas University: Duzce University - Computer Engineering
CrownCode is a web-based platform hosting multiple open-source projects ranging from AI-powered audio analysis to creative tools. Each project is built with modern technologies and designed for real-world use.
AURIS uses a Multi-Tower Detection Architecture to distinguish AI-generated music from human-composed tracks:
Audio Input
|
+---> Tower 1: wav2vec2-base (deep audio embeddings)
+---> Tower 2: 49 Handcrafted Features (spectral, temporal, harmonic, vocal)
+---> Tower 3: CLAP (cross-modal audio-text similarity)
+---> Tower 4: FST API (external spectral fakeprint detection)
|
v
Meta-Classifier (RF / XGBoost / LightGBM / SVM / MLP)
|
v
Final Prediction: AI-Generated vs Human-Composed
| Project | Description | Status |
|---|---|---|
| AURIS AI Music Detection | Multi-tower AI/human music classifier | Active |
| ML Toolkit | Data augmentation & processing tools | In Development |
| Crown Fortune | Interactive fortune wheel game | Active |
| Crown Dreams | Dream interpretation with AI | Active |
| Crown Commend | YouTube comment generator | Active |
| Crown Vote | Voting & polling system | In Development |
CrownCode/
├── platform/ # Next.js 14 Frontend (Pages Router)
│ ├── pages/ # Route pages (14 routes)
│ ├── components/ # React components
│ │ ├── Layout/ # MainLayout, Header, Footer
│ │ ├── Home/ # HeroSection, ProjectsSection
│ │ ├── AurisDetection/ # HeroSection, HowItWorks, AnalysisResultCard
│ │ ├── Navigation/ # HeaderNavbar, SearchModal
│ │ └── UI/ # Toast, Skeleton, CopyButton
│ ├── styles/ # CSS Modules + global design system
│ │ ├── base/ # variables.css, reset, typography, animations
│ │ ├── components/ # Component-specific CSS
│ │ └── pages/ # Page-specific CSS modules
│ ├── context/ # LanguageContext, ToastContext
│ ├── hooks/ # useFileAnalysis, useYouTubeAnalysis, etc.
│ ├── locales/ # en.json, tr.json (full i18n)
│ ├── config/ # product-catalog.ts
│ └── public/ # Static assets, images, icons
├── backend/ # FastAPI Backend
├── hf-crowncode-backend/ # HuggingFace Spaces deployment
├── DataSet/ # AURIS training data & download scripts
│ ├── download_datasets.py # Streaming dataset downloader
│ └── metadata.csv # Master manifest
├── Android-App-CrownCode/ # Android native app
├── docs/ # Documentation
│ ├── academic/ # Thesis reports (TR/EN)
│ ├── technical/ # AI_MODEL_STRATEGY, architecture docs
│ ├── guides/ # Quick start, code examples
│ ├── community/ # Code of conduct, contributing
│ └── notes/ # Analysis reports
├── scripts/ # Utility scripts
└── netlify/ # Deployment config
- Framework: Next.js 14 (Pages Router) + React 18 + TypeScript
- Styling: CSS Modules + custom design token system (variables.css)
- Animation: Framer Motion 11
- Icons: Lucide React
- i18n: Custom LanguageContext (Turkish / English)
- Deployment: Netlify
- Runtime: Python 3.11+
- Framework: FastAPI (deployed on HuggingFace Spaces)
- AI Models: wav2vec2-base, CLAP (HTSAT-base), librosa feature extraction
- ML: PyTorch, scikit-learn, XGBoost, LightGBM
- Audio Processing: librosa, torchaudio, soundfile
- Frontend Hosting: Netlify (automatic builds)
- Backend Hosting: HuggingFace Spaces
- SEO: JSON-LD structured data (Organization, WebSite, SoftwareApplication, BreadcrumbList)
- PWA: manifest.json, service worker ready
| Route | Description |
|---|---|
/ |
Homepage - Hero + Projects showcase |
/ai-music-detection |
AURIS detection tool + tech stack |
/data-manipulation |
ML Toolkit interface |
/crown-fortune |
Fortune wheel game |
/crown-dreams |
Dream interpretation |
/crown-commend |
YouTube comment generator |
/crown-vote |
Voting system |
/creator-studio |
Creator tools |
/analysis-history |
Past analysis results |
/system-status |
System health dashboard |
/search |
Global search |
/privacy |
Privacy policy |
/terms |
Terms of service |
/api/health |
Health check endpoint |
/api/version |
Version info endpoint |
- Node.js 20+
- Python 3.11+ (for backend & dataset tools)
# Clone
git clone https://github.com/Rtur2003/CrownCode.git
cd CrownCode
# Install frontend dependencies
cd platform
npm install
# Run development server
npm run dev
# -> http://localhost:3000
# Production build
npm run buildcd DataSet
pip install datasets soundfile numpy
python download_datasets.py --resumeTraining data sourced from HuggingFace via streaming download:
| Source | Type | Target |
|---|---|---|
| SleepyJesse/ai_music_large | AI + Human | 4,000 |
| disco-eth/AIME | AI (12 models) | 1,000 |
| marsyas/gtzan | Human (10 genres) | 1,000 |
| benjamin-paine/free-music-archive-small | Human | 1,000 |
| zuhri025/suno-audio | AI (Suno) | 500 |
| UniDataPro/real-vs-fake | Vocal deepfake | 1,000 |
All audio resampled to 16kHz mono WAV, trimmed to 30s max.
See DataSet/README.md for full details.
- AI Model Strategy - Multi-tower architecture details
- Modular Architecture
- Technology Stack
- Deployment Config
- Read Development Guidelines
- Create a topic branch from
gelistirme - Make atomic commits using conventional commit format
- Submit a pull request
See CONTRIBUTING.md for full details.
MIT License - see LICENSE
@thesis{altuntas2026crowncode,
title={Web-Based AI Music Detection and Data Manipulation Platform},
author={Hasan Arthur Altuntas},
institution={Duzce University},
department={Computer Engineering},
year={2026},
type={Bachelor's Thesis},
url={https://hasanarthuraltuntas.xyz}
}- GitHub: @Rtur2003
- Website: hasanarthuraltuntas.xyz
- Email: [email protected]