CUVIS.AI Documentation

CUVIS.AI is a modular, low-code/no-code framework for building reproducible machine-learning pipelines for hyperspectral data analysis. It provides a thin abstraction over PyTorch, PyTorch Lightning, and Hydra, with reusable nodes you can compose into graph-based HSI workflows.

What you can do

• Typed I/O System: Port-based connections with type safety and validation
• Statistical Initialization: Bootstrap models with non-parametric methods (RX detector, PCA)
• Gradient-Based Training: Fine-tune models with PyTorch Lightning
• Flexible Node Architecture: Composable preprocessing, feature extraction, and decision modules
• Comprehensive Monitoring: Integrated TensorBoard support and extendible to other frameworks
• Configuration Management: Hydra-based configuration with CLI overrides

Quick Links

• Installation
• Quickstart
• Core Concepts
• API Reference
• Tutorials

Ready to get started?

• Start with the Installation Guide, then follow the Quickstart.
• Want to contribute? See the Contributing Guide.
• Found an issue? Report bugs / request features

---

Apache License 2.0 — see LICENSE.