una-dinosauria / 3d-pose-baseline

A simple baseline for 3d human pose estimation in tensorflow. Presented at ICCV 17.
MIT License
1.42k stars 353 forks source link

Focal length has 2 values #167

Open ehedlin opened 4 years ago

ehedlin commented 4 years ago

Hey, I have a question about cameras.h5. Each camera seems to have 2 focal lengths (shape 2x1). All other entries seem to make sense to me (translation is a 3x1, rotation matrix is a 3x3x1, etc). Which is the right value or is there something I am missing?

una-dinosauria commented 4 years ago

Great catch! Apparently our documentation is wrong 😅

They are both right. Often the x and y components of the focal length can differ due to a number of reasons; eg from http://ksimek.github.io/2013/08/13/intrinsic/

  • Flaws in the digital camera sensor.
  • The image has been non-uniformly scaled in post-processing.
  • The camera's lens introduces unintentional distortion.
  • The camera uses an anamorphic format, where the lens compresses a widescreen scene into a standard-sized sensor.
  • Errors in camera calibration.

I'll be very thankful if you submit a PR correcting the documentation.