Meta's Segment Anything (SAMIST)

Segment Anything Model (SAM) produces high quality object masks from input prompts such as points or boxes, and it can be used to generate masks for all objects in an image. - SAM.cpp / HN

SAM.cpp

Needs pytorch for model conversion

$ podman run --gpus all --rm -ti --ipc=host -v /home/yves/DEV/Segmentatio/sam.cpp/:/workspace pytorch/pytorch:latest
# Convert PTH model to ggml. Requires python3, torch and numpy
python convert-pth-to-ggml.py checkpoints/sam_vit_b_01ec64.pth . 1
# run inference
$ ./bin/sam -t 16 -m ../checkpoints/ggml-model-f16.bin -i ../img.jpg

SAMIST

list different models with different size.

caption

Written on September 6, 2023, Last update on February 16, 2024
segmentation AI project