search

haofengac / MonoDepth-FPN-PyTorch

Single Image Depth Estimation with Feature Pyramid Network
MIT License
326 stars 69 forks source link

Problem with choosing ground truth form, filled depth or project depth #9

Open tjqansthd opened 5 years ago

tjqansthd commented 5 years ago

Hi @xanderchf

Thanks to your post, I tried to fill in missing values in ground truth by using NYU toolbox - fill_depth_colorization in matlab. and applied colorized map, g t_1 like this.

But I'm curious about one thing, when you compute errors between ground truth and prediction, did you use filled depth or project depth? In other words, is your network outputs filled depth or project depth? Thanks