I am Mikita Misiura, an astrophotographer and a chemist turned data scientist turned software developer. StarNet and DeepSNR both started as tools I wanted for my own processing workflow: first to make star removal more practical, and later to explore denoising approaches that worked better for the kind of astronomical data I was processing.
Neither project started as some kind of a product. The tools evolved as I learned more about machine learning, software engineering, packaging, and the practical details of getting reliable software into other astrophotographers' workflows. After StarNet was released publicly, it found an audience in the astrophotography community and became part of many PixInsight and command-line workflows. DeepSNR follows a similar path for noise reduction.
This website is the new shared home for that work. It brings together the old StarNet and DeepSNR websites so installation instructions, downloads, documentation, release notes, and FAQ entries can live in one place and stay easier to maintain.
The tools are free, and the project is shaped by real-world use: my own processing, workflow experiments, and the occasional third-party wrapper or integration built by other astrophotographers.
