search

Brummi / MonoRec

Official implementation of the paper: MonoRec: Semi-Supervised Dense Reconstruction in Dynamic Environments from a Single Moving Camera (CVPR 2021)
MIT License
587 stars 85 forks source link

Clarification regarding mask module #10

Closed testingshanu closed 3 years ago

testingshanu commented 3 years ago

I wanted to clarify the following:

  1. What type of information does stereo image add to the monorec network?
  2. Is it possible to train the end to end model only using monocular images ?
  3. Is it possible to train only the mask module only using monocular images ?
nynyg commented 3 years ago

I will try to answer your questions in the following:

  1. The stereo images make the depth training more accurate, are crucial for the mask module refinement, and are needed for generating the auxiliary masks.
  2. For the depth module, yes, for the mask module, as I mentioned above, we need the stereo images for the mask module refinement and generating auxiliary masks.
  3. Please refer to 2.

Thanks