YOLOV8_SAM

yolov8 model with SAM meta

Use yolov8 & SAM model to get segmention for custom model

installation

pip install ultralytics
pip install 'git+https://github.com/facebookresearch/segment-anything.git'

Download weights

 !wget -P images https://raw.githubusercontent.com/facebookresearch/segment-anything/main/notebooks/images/truck.jpg
 !wget -P images https://raw.githubusercontent.com/facebookresearch/segment-anything/main/notebooks/images/groceries.jpg       
 !wget https://dl.fbaipublicfiles.com/segment_anything/sam_vit_h_4b8939.pth

Test on image

python detect_multiple_object_SAM.py

visulise the results

python3 visulise_mask.py

Results

Bounding box: [478, 1280, 182, 76]

Segmentation mask: [631, 1280, 630, 1281, 629, 1281, 628, 1282, 626, 1282, 625, 1283, 622, 1283, 621, 1284, 619, 1284, 618, 1285, 615, 1285, 614, 1286, 612, 1286, 611, 1287, 609, 1287, 608, 1288, 607, 1288, 606, 1289, 604, 1289, 603, 1290, 602, 1290, 601, 1291, 599, 1291, 598, 1292, 596, 1292, 595, 1293, 593, 1293, 592, 1294]

Save the result in yolo format for training Mask segmentation model.

yolo format = [0 0.529687 0 0.014815 0 0.529167 0 0.015741 0 0.525521 0 0.015741 0 0.525000 0 0.016667 0 0.519792 0 0.016667 0 0.519271 0 0.017593 0 0.513021 0 0.017593 0 0.512500 0 0.018519 0 0.505208 0 0.018519]

TODO

- Doing annotations on multiple images  - Done
- Add support for saving annotations in yolo format -Done
- Support jsno format for segmentation model trainig

refrence

https://github.com/facebookresearch/segment-anything
https://github.com/ultralytics/ultralytics