search

Eli-YiLi / PMM

Pseudo-mask Matters in Weakly-supervised Semantic Segmentation
MIT License
44 stars 4 forks source link

About args.gen_seg_mask #10

Closed ZechengLi19 closed 2 years ago

ZechengLi19 commented 2 years ago

When I generate mask on COCO2017. It seems need to a lot of time to finish this work. I use 8 T4 or 2 RTX3090 to test the speed. Both of them get 200 .npy files for an hour. It means I need nearly 11 days to generate mask. Could you tell me. What can I do for speed up this code？Thanks.

Eli-YiLi commented 2 years ago

You can reduce the number of scales in multiscale test, especially high resize scale. But the performance is not promised, you'd better check the performance on a tiny subset. Note that I ran it on 4 x 8 1080Ti (similar to 8 3090Ti).

ZechengLi19 commented 2 years ago

Thanks! And I want to know, how long did you spend for generating mask?