Open mpuu00001 opened 1 year ago
Hi, based on your question, I suggest that you could customize a data loader for your own dataset. The coordinates of each hand landmark can be obtained from the networks prediction, see the jointsRS. (RS: Root relative and Scale normalized)
Thank you for the suggestion! To make my life easier, alternatively, is there a way for me to run the model onto a single image instead of a whole dataset?
Hi! I want to apply the model onto a different dataset other than RHD and STB, then get the coordinates of each hand landmark point for each hand presented.
May I know how can I do that?