search

YudeWang / SEAM

Self-supervised Equivariant Attention Mechanism for Weakly Supervised Semantic Segmentation, CVPR 2020 (Oral)
MIT License
539 stars 97 forks source link

CAM is not accurate. #8

Closed serend1p1ty closed 4 years ago

serend1p1ty commented 4 years ago

Hi, I know the performance of weakly-supervised semantic segmentation is not so well as supervised SS.

But I still confused about the result, it's much worse than I thought.

Original image: image

CAM: image

I just run this command: python infer_SEAM.py --weights ../resnet38_SEAM.pth --infer_list voc12/val.txt --out_cam_pred out_cam_pred

JingyiWang2021 commented 2 years ago

Hi @serend1p1ty ,

Did you find the answer of your question? I am trying to use SEAM to generate CAMs of another dataset, but I don't know how to do it. Do you think that I can make it by just running this command: python infer_SEAM.py --weights ../resnet38_SEAM.pth --infer_list voc12/val.txt --out_cam_pred out_cam_pred ?