Input data interpretation

[engrjavediqbal]

Hi, thank you for releasing the source code for such great work. I am new in this field and please accept my apologies for asking very basic questions.

You told in issue #6, to run DeMoN to get poses_demon.txt and respective depth maps. However, it seems that the DeMoN evaluation (for image-1 as reference and image-2 as source images ) only outputs the depth map for the first image and the relative pose for the second image. Can you please elaborate on how to generate the depth map and pose for the second image? e.g. by simply swapping the input image pair?

Second, in the provided examples (poses.txt), there are two poses there, can you please elaborate which pose corresponds to which image, i.e., line-1 in poses.txt correspond to the relative pose of image-1 with reference to image-2?

Third, the translation values in the poses.txt seem to be normalized between (0,1), which in the ground truth of MVS dataset are in higher values. Can you please elaborate on how they are normalized? or what pre-processing is used?

Thank you.

[weixk2015]

Q1. Yes, swapping the image pair could generate the depth map for the second image.
Q2. No, each pose is an absolute pose. To cal the rel pose from source img-1 to ref img-2, try pose1.matmul(pose2.inverse)
Q3. The data from MVS dataset is pre-processed and provided by DeMoN.

[engrjavediqbal]

@weixk2015 Thank you for your reply. I have one more question. The DeMoN (as you answered in Q1 above) will provide relative poses for img-1 and img-2 with reference to each other. However, you said in Q2, that the poses are absolute poses (with reference to origin). Can you please elaborate further the input poses to DeepSFM, i.e., are the input poses to deepSFM are absolute (with reference to origin)? if yes, then how are they made absolute once obtained from DeMoN.

Second, if the input poses to deepSFM are relative, are they the same as produced by DeMoN or any type of conversion or processing is applied.

Thank you.

[wangxiyuan9]

@engrjavediqbal @weixk2015 hello， could you Reproduce the experimental results of this paper ，My results are very different from the paper. After inspection , it seems because the initial posture data from demon is quite different from pose.txt. I don't know why . These initial poses are obtained by running the example.py file in the demon with given two photos. Can you share how you obtained the initial poses from the demon and the level of difference between them and pose.txt? thanks