AI image segmentation

Sam is an image segmentation model implemented from scratch in C++. It can perform pixel-level segmentation of images and locate object boundaries without any additional code and annotations. Based on Meta's Segment Anything Model, Sam uses the Transformer architecture for end-to-end image segmentation prediction. It provides a simple and easy-to-use C++ interface, supporting both command line and graphical interface usage. Sam can run efficiently on the CPU, the model is compact, and it also ensures good segmentation accuracy. It is ideal for deploying and using image segmentation models in embedded environments where high performance is required but GPUs are not available.