Instructions to use IDEA-Research/grounding-dino-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use IDEA-Research/grounding-dino-base with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("zero-shot-object-detection", model="IDEA-Research/grounding-dino-base")# Load model directly from transformers import AutoProcessor, AutoModelForZeroShotObjectDetection processor = AutoProcessor.from_pretrained("IDEA-Research/grounding-dino-base") model = AutoModelForZeroShotObjectDetection.from_pretrained("IDEA-Research/grounding-dino-base") - Notebooks
- Google Colab
- Kaggle
Bounding box width and height too large
#13
by jmPearl - opened
It seems that the bounding box for this model is inaccurate. The x,y are spot on, but across many images I've found the width and height to be far too large. Sometimes the bounding box even goes off the size of the image. The model works great locally not using HF, but when using the HF version it does not work as expected.
jmPearl changed discussion status to closed
I was. It was user error. I mistook the third and fourth elements in the box to be width and height. I now know they are xmax and ymax.