FastSAM
FastSAM copied to clipboard
Doubts about the box detection result, whether it is correct?
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?