search

shariqfarooq123 / AdaBins

Official implementation of Adabins: Depth Estimation using adaptive bins
GNU General Public License v3.0
727 stars 156 forks source link

I want to train my dataset #23

Closed Seonguke closed 3 years ago

Seonguke commented 3 years ago

Hi As the title, I have some questions about train for my dataset.

  1. what does 'focal' mean & How does that affect learning?
  2. How much loss is required for a good output?
shariqfarooq123 commented 3 years ago

Hi

As the title, I have some questions about train for my dataset.

  1. what does 'focal' mean & How does that affect learning?

  2. How much loss is required for a good output?

  1. 'focal' values are the camera params and are not used in this paper.
  2. Your second question is very difficult to answer even in the whole context of deep learning. However, you can keep a validation set. Then lower the validation loss, the better.