NISPS Playground
Train a neural network to map joystick positions to output parameters using interactive machine learning.
Modes
Open the side panel (☰) to switch between Visual and Synth modes.
Visual Mode
20 outputs control a flow-field particle system.
Synth Mode
20 outputs control a C15 synthesizer engine. Use the arpeggiator for automatic note playback.
Examples Mode
- Move the joystick to a position
- Drag the parameter bars to set desired output
- Press Add Example to save this mapping
- Repeat for different joystick positions
- Press Train — the network learns to interpolate between your examples
RL Feedback Mode
- Move the joystick — the network produces outputs
- See something you like? Press + (thumbs up)
- Don't like it? Press − (thumbs down) to explore more
- Double-click the joystick to toggle Follow mode
Tips
Your training data auto-saves to your browser. Gamepad supported (LB/RB for RL feedback).
Tap anywhere outside to close