Looking for a Model or Dataset for Detecting Objects Held in Hand

[YoungjaeDev]

Hi,

I’m trying to detect objects held in hand. Do you know of any models or datasets that are well-suited for this task?

If labeling is required, would it be better to use YOLO-world for bbox grounding?

Additionally, there are a large number of product classes involved. I’m wondering if it would be better to only detect objects and handle class recognition through retrieval methods.

Thank you!