I mean, after reviewing all the open and closed issues, it appears as though you're purposefully avoiding providing users with the ability to run inference on an RGB image+corresponding single channel mask to obtain a refined mask, as the title of your repository advertises. The license of your model is Apache, which indicates a very broad freedom to use your model. The problem is,
your repository does not clearly and simply instruct a user to conduct inference.
...or is it that your model doesn't work for in-the-wild inference...? In which case, one is lead to believe that there's another model that you have created that indeed works for in-the-wild inference which, for one reason or another, you've chosen to withhold from the public.
So, what's up? This is the third issue I've opened regarding running inference.
I mean, after reviewing all the open and closed issues, it appears as though you're purposefully avoiding providing users with the ability to run inference on an RGB image+corresponding single channel mask to obtain a refined mask, as the title of your repository advertises. The license of your model is Apache, which indicates a very broad freedom to use your model. The problem is,
your repository does not clearly and simply instruct a user to conduct inference.
...or is it that your model doesn't work for in-the-wild inference...? In which case, one is lead to believe that there's another model that you have created that indeed works for in-the-wild inference which, for one reason or another, you've chosen to withhold from the public.
So, what's up? This is the third issue I've opened regarding running inference.