The point is that the model works fast but bpd_cuda.forward call takes a large amount of time so the question is:
what output data is and can I not to call bpd_cuda.forward to get data about segments?
Also is the network orientied for specific domain image content?
Cause for my test images it results bad.
The point is that the model works fast but
bpd_cuda.forward
call takes a large amount of time so the question is: what output data is and can I not to callbpd_cuda.forward
to get data about segments?Also is the network orientied for specific domain image content? Cause for my test images it results bad.
Original image![716c971s-960](https://user-images.githubusercontent.com/51739187/112950899-782d4080-9143-11eb-8196-59f31315180a.jpg)
Visualization grid![result](https://user-images.githubusercontent.com/51739187/112950987-8f6c2e00-9143-11eb-9dbd-d72ac4080589.png)
root![root](https://user-images.githubusercontent.com/51739187/112951077-a743b200-9143-11eb-9e69-b8bc0eda9f80.png)
super_BPDs![super_BPDs](https://user-images.githubusercontent.com/51739187/112951117-b165b080-9143-11eb-9057-c1e2cf15c0d4.png)
super_BPDs_before_dilation![super_BPDs_before_dilation](https://user-images.githubusercontent.com/51739187/112951166-bdea0900-9143-11eb-8827-60e61215fe8d.png)
super_BPDs_after_dilation![super_BPDs_after_dilation](https://user-images.githubusercontent.com/51739187/112951187-c3475380-9143-11eb-924c-da302825a77e.png)
my code to inference single image: