Zach Teed

Tensorflow 1.12.0 should have custom_gradient. What happens when you run these commands? ``` >>> import tensorflow as tf >>> tf.__version__ '1.12.0' >>> tf.custom_gradient ```

Hi, I used the split used in the BA-Net paper in order to compare to BA-Net. The images/depths/poses were extracted from the .sens file with frame skip = 1. I...

I will post a .txt file on the cases where DSO succeeds later today or tomorrow. I have the logs from this experiment arxived, but I will need to parse...

These are the poses I got from running dso [dso_poses.zip](https://github.com/princeton-vl/DeepV2D/files/6094524/dso_poses.zip) You can use this script to parse the results, you should find that dso has the poses for 1665 of...

Yeah that doesn’t look good, thanks for bringing this up. It looks like there is some bias in the driving scenarios causing the scale to constantly expand. I will look...

If you run with the `--mode=global` the depth maps for all frames will be predicted. You can also try the SLAM demos for predicting depth for longer sequences

No that is not intended. But these are floating point values so there won't be any real difference in behavior between > and >=

Hi Azin, I did not upload results from mixed precision training to the Sintel leaderboard. I only evaluated the mixed precision training on KITTI and Sintel train splits after training...

Hi, you need to compile the correlation kernel with the same major cuda version which PyTorch was built with. PyTorch 1.6 was built with CUDA 10, so you would need...

Hi, thanks for bringing this to my attention. There appears to be some issue with loading the model weights. I am looking into this today and will post a corrected...