bowang-lab/U-Mamba

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1019 stars · 108 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

U-Mamba is a hybrid CNN-state-space-model (Mamba) network for biomedical image segmentation, built on top of the nnU-Net framework. It combines convolutional local feature extraction with Mamba's long-range dependency modeling and self-configures to datasets for 2D and 3D segmentation tasks.

Use cases

• segment organs in CT and MRI scans
• segment cells in microscopy images
• segment surgical instruments in endoscopy video
• train a segmentation model on my own medical dataset
• try Mamba-based architectures for medical image segmentation
• get state-of-the-art biomedical segmentation without manual tuning

When to choose

• you need long-range dependency modeling in 2D/3D medical image segmentation
• you already use nnU-Net and want a stronger network architecture
• you work with CT, MRI, endoscopy, or microscopy segmentation tasks

When to avoid

• you need a lightweight CPU-only solution - it requires CUDA GPUs and Mamba SSM kernels
• your task is general natural-image segmentation rather than biomedical
• you need Windows or macOS support - it targets Ubuntu with CUDA 11.8

Facets

library · maturity active

machine-learning deep-learning image-processing computer-vision deep-learning computer-vision healthcare bioinformatics python image-segmentation mamba state-space-models nnunet biomedical-imaging medical-imaging research-code linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="bowang-lab/U-Mamba")