facebookresearch / consistent_depth

We estimate dense, flicker-free, geometrically consistent depth from monocular video, for example hand-held cell phone video.
MIT License
1.63k stars 236 forks source link

What does the ParameterLoss mean? #18

Closed XYZ-qiyh closed 4 years ago

XYZ-qiyh commented 4 years ago

Hello , thanks for your amazing work! I wonder to know what does the ParameterLoss mean. In your paper, there are only spatial loss and disparity loss (ConsistencyLoss in this repo). Looking forward to your reply.

XYZ-qiyh commented 4 years ago

I got it! The lambda_parameter was set to 0 in this repo. The ParameterLoss is not used here.