IDEA-Research/Grounded-SAM-2

SAM 2 Florence-2 pipeline image

spped2000 opened this issue · 3 comments

Do you have an image of the Grounded SAM 2 + Florence-2 pipeline, similar to the Grounded SAM 2?

Do you have an image of the Grounded SAM 2 + Florence-2 pipeline, similar to the Grounded SAM 2?

Hi @spped2000 , You mean the visualization results using Florence-2 + SAM 2 or the demo image? You can try the following script to get the visualization results:

python grounded_sam2_florence2_image_demo.py \
    --pipeline object_detection_segmentation \
    --image_path ./notebooks/images/cars.jpg

We name it grounded_sam2_florence2_image_demo because it's under our Grounded SAM 2 series, the algorithms behind this are Florence-2 + SAM 2

Do you have an image of the Grounded SAM 2 + Florence-2 pipeline, similar to the Grounded SAM 2?

Hi @spped2000 , You mean the visualization results using Florence-2 + SAM 2 or the demo image? You can try the following script to get the visualization results:

python grounded_sam2_florence2_image_demo.py \
    --pipeline object_detection_segmentation \
    --image_path ./notebooks/images/cars.jpg

We name it grounded_sam2_florence2_image_demo because it's under our Grounded SAM 2 series, the algorithms behind this are Florence-2 + SAM 2

I mean something like this
image

Do you have an image of the Grounded SAM 2 + Florence-2 pipeline, similar to the Grounded SAM 2?

Hi @spped2000 , You mean the visualization results using Florence-2 + SAM 2 or the demo image? You can try the following script to get the visualization results:

python grounded_sam2_florence2_image_demo.py \
    --pipeline object_detection_segmentation \
    --image_path ./notebooks/images/cars.jpg

We name it grounded_sam2_florence2_image_demo because it's under our Grounded SAM 2 series, the algorithms behind this are Florence-2 + SAM 2

I mean something like this image

Yes, it's the same pipeline, but using Florence-2 instead of Grounding DINO