How to reproduce Agentic Object Detection results using this library?

[bittergourd1224]

I have tried some image on the demo website https://va.landing.ai/demo/agentic-od, the results look pretty good.
But when I tried to reproduce the results using this library, with function "countgd_object_detection" according to README, the results are not good as expected.
Any advices? Thanks.

[dillonalaird]

You can access AgenticOD through the tools right now, I just have not tested it with the VisionAgent yet (so VisionAgent cannot see the AgenticOD tool, but it is available to users). Here's a snippet of how to use it:

import vision_agent.tools as T
import matplotlib.pyplot as plt
image = T.load_image("your_image.png")
dets = T.agentic_object_detection("person wearing helmet", image)
viz = T.overlay_bounding_boxes(image, dets)
plt.imshow(viz)
plt.show()