Can we fine tune this model for specific use case?

[sid8491]

I have a use case where I need to remove the background from the pictures of laptops.
In some pics, a person is holding a laptop in hands, in some, there is a white background, etc.
Can I fine-tune the model for this?
Using it directly is not giving perfect results.