good day and congratulations for your great work,
I have been testing it with more generic human pictures (like groups of people with overlapping of limbs), and the results are not very good (using your pretrained model that you provided). Is this because your datasets are focused on just portrait pics of single people?
to work with these kind of human group pictures do you think that I would need to retrain the model with something like COCO or similar?
good day and congratulations for your great work, I have been testing it with more generic human pictures (like groups of people with overlapping of limbs), and the results are not very good (using your pretrained model that you provided). Is this because your datasets are focused on just portrait pics of single people? to work with these kind of human group pictures do you think that I would need to retrain the model with something like COCO or similar?
thank you for any tips :)