11 repos
Interactive and automated image segmentation techniques, centered on the Segment Anything Model (SAM) and its variants for real-time, prompt-driven segmentation across diverse visual domains. Repositories here cover core SAM implementations, performance-optimized versions like FastSAM, robustness improvements, multimodal extensions that integrate vision with language models, and applications like referring image segmentation. Most materials are in Python and Jupyter Notebook, making this a practical learning resource for both understanding segmentation foundations and deploying efficient models in production.