Change the size of the input image

zhengqili / MegaDepth

Code of single-view depth prediction algorithm on Internet Photos described in "MegaDepth: Learning Single-View Depth Prediction from Internet Photos, Z. Li and N. Snavely, CVPR 2018".

MIT License

723 stars 169 forks source link

Change the size of the input image #44

Closed yanyu-holo closed 3 years ago

yanyu-holo commented 3 years ago

Hello, your work is very helpful to me. I want the input RGB image and the output depth map to be 1000 * 1000 pixels. How can I change the code?

zhengqili commented 3 years ago

our current method only supports the size of multiple of 64，you might need to resize the image before feeding the network and resize predicted depth back to 1000x1000

yanyu-holo commented 3 years ago

Thank you for your reply. If I understand correctly, your method is to get relative depth, not absolute depth, right?

zhengqili commented 3 years ago

Hi, yes. Only relative depth, up to scale.