Fangchang Ma

Results 28 comments of Fangchang Ma

Hey Chris, Thanks for point this out -- the comment was outdated, and we are indeed doing bottom crop (same as the KITTI dataset).

> Are there any other locations in the code where the intrinsics are used? There is an Intrinsics class in inverse_warp.py that performs similarly to your code. https://github.com/fangchangma/self-supervised-depth-completion/blob/898aa91f569adb1165fee82d4318025fae78546a/inverse_warp.py#L5 If you...

The downsampling process is non-trivial. We kept a subset of the laser scans, which were then projected onto the image plane. However, since some lidar measurements were missing in the...

Please provide more details.

@gautamjain1009 Thanks for the screenshot. It turns out that the problem was due to a version difference in the code and the previously saved results. I just pushed a minor...

By default we use the `../data/kitti_depth/val_selection_cropped` folder for validation. This is the "[manually selected validation and test data sets (5 GB)](http://www.cvlibs.net/download.php?file=data_depth_selection.zip)" on the KITTI website. I am not entirely sure...

It seems that the file structure of KITTI Depth has changed recently. I will update my code accordingly. Please stay tuned.

Just updated the code and the README for the latest file structure.