In a notebook

I have three thorough tutorials on using Omnipose in a Jupyter notebook. The first focuses on mono-channel segmentation of locally saved images. The second shows how to use a second channel to aid in segmentation, also taking the opportunity to demonstrate how to download images on-the-fly in a notebook. The third shows how to do 3D segmentation, dealing with GPU VRAM bottlenecks and visualization strategies along the way.

• The basics in 2D
• Multiple channels
• Omnipose in 3D

See settings for more information on algorithm parameters.