🍺 SkufUp — Beer Can Opening Detector

SkufUp listens to your microphone and waits for the characteristic "pshhh" sound of opening a beer can.
When it hears it — automatically launches a game or opens a website!

Before beer 😢		After beer 🍺
	➡️

• 🤖 ML Detector — trained on real recordings, ~95% accuracy
• 🎮 Process Check — won't open the game if it's already running
• 🌐 Website Support — can open any URL
• 🚀 Autostart — launches with Windows
• 🎨 Beautiful Interface — dark theme, minimalist design

Download SkufUp_Setup.exe from Releases and run it.

# Clone the repository
git clone https://github.com/your-username/SkufUp.git
cd SkufUp

# Install dependencies
pip install -r requirements.txt

# Run
python gui_app.py

1. Launch SkufUp
2. Choose a game (.exe) or specify a website
3. Press START
4. Open a beer 🍺
5. Enjoy! 🎮

🎤 Microphone → 📊 Audio Analysis → 🤖 ML Model → ✅ Beer!

Characteristic features of "pshhh":

• 📈 High frequencies (2-8 kHz) — gas hissing
• ⚡ Fast attack — sudden onset
• 📉 Short duration — 100-500 ms

Source Code/
├── gui_app.py              # 🖥️ Main application with GUI
├── audio_detector_ml.py    # 🤖 ML sound detector
├── audio_detector.py       # 📊 Basic detector (fallback)
├── train_model.py          # 🎓 ML model training
├── analyze_audio.py        # 📈 Audio file analysis
├── config.py               # ⚙️ Settings
├── beer_sound_template.py  # 🍺 Reference sounds
├── build_installer.py      # 📦 Installer builder
└── requirements.txt        # 📋 Dependencies

Open config.py:

DETECTOR_SETTINGS = {
    "similarity_threshold": 0.45,  # Similarity threshold
    "peak_threshold": 0.1,         # Volume threshold
    "cooldown": 3.0,               # Pause between triggers
    "debug_mode": True,            # Show logs
}

# Build application
pyinstaller SkufUp.spec --noconfirm

# Create installer
python build_installer.py
pyinstaller --onefile --windowed --name=SkufUp_Setup --icon=setup.ico SkufUp_Installer.py

If you want to improve accuracy:

# 1. Put your recordings in sound/ folder
# 2. Analyze
python analyze_audio.py

# 3. Train model
python train_model.py

# 4. Rebuild exe
pyinstaller SkufUp.spec --noconfirm

Enable debug_mode: True and watch the console:

🔈 [████████░░░░░░░░░░░░] RMS:0.023 Peak:0.089
🔊 [████████████████████] RMS:0.156 Peak:0.423
   🤖 ML probability: 89% | H/L: 15.2 | Centroid: 8500Hz
🍺 CAN OPENED! (ML: 89%, H/L: 15.2)