JingZhang617/Scribble_Saliency
Weakly-Supervised Salient Object Detection via Scribble Annotations, CVPR2020
Python
Issues
- 0
After adding the edge information, isn't this a fully supervision, why is it still weakly supervision?
#17 opened by yike5022 - 0
We have scribble dataset (1: foreground, 2: background, 0: unknown), raw RGB images, gray images and edge map from:https://github.com/yun-liu/rcf.
#16 opened by yike5022 - 0
May I ask whether to use the Scribble image directly with the image as the loss when calculating the loss, or to convert the Scribble to a black and white gt image before making the loss?
#15 opened by yike5022 - 0
Could you provide more details on the image intensity value generation logic?
#14 opened by JanineCHEN - 2
- 2
[Errno 2] No such file or directory: '/home/jing-zhang/jing_file/vgg16-397923af.pth'
#13 opened by 18718277909 - 0
- 0
Could you provide results of other models?
#10 opened by JZTgentle - 2
- 0
How do you get the edge map by sobel filter?
#8 opened by lzyhha - 11
About the smoothness loss
#3 opened by lianxxx - 0
ModuleNotFoundError: No module named 'vgg'
#7 opened by andy500 - 0
Could you provide the metric evaluation code?
#6 opened by tjulyz - 1
- 2
about the uploaded trained model
#4 opened by siyueyu - 6
When are you going to release the code?
#1 opened by kayleeliyx - 0
Evaluation Metric
#2 opened by GrassBro