Open githublqs opened 1 year ago
This is the final output of the three masks
Hello, is the problem the third mask?
@zxDeepDiver yes
OK, thanks for yout feedback. Currently FastSAM indeed generates these masks for some images, as discussed in our paper. We are working to improve it.
This is using c++?
Can you share your codes?