search

nqanh / affordance-net

AffordanceNet - Multiclass Instance Segmentation Framework - ICRA 2018
Other
121 stars 54 forks source link

about results #13

Closed lsj910128 closed 6 years ago

lsj910128 commented 6 years ago

Hi, do you know why the segmented result (the area of the target object) is always smaller than the ground truth? Someone told me the reason is threshold for the segmentation mask in the method is too large. But I don't what is this threshold?

lsj910128 commented 6 years ago

Hi, I have a question about Robust Resizing Affordance Mask: how to set the hyperparameter α and where to set? How can I make α learned during the network training phase.

nqanh commented 6 years ago

During training, alpha is 0.005 (line 478 and 485 in proposal_target_layer.py). During testing, you can change line 20 in demo_img.py.

This param is not learnable.

lsj910128 commented 6 years ago

Thank you for your reply. But how to determine the value of alpha?

nqanh commented 6 years ago

It's a hyperparameter. Setting a higher value of alpha will lead to noisier affordance map. You may want to play with this param in the demo script.