chamorajg / pytorch_depth_and_motion_planning Goto Github PK

Thank you for this work,
How can I display the depth map images using this code?

Thank you for your gorgeous work. I am trying your project and I've get the position.txt and trajectory.txt. But what should I do to see exactly the depth map of an image format?

Thank you for this work,
I want to use this approach for camera self calibration. I trained the network on KITTI dataset and used the model to test inference. However, the obtained intrinsic_mat (test.py) is very diffrent from Kitti ground thruth calibration matrix.
Could you give an explanation to this?

This is a great project and I would like to experiment with it.

Could you please add a license for this project, otherwise one isn't allowed to use it for anything?

Hello,

Thanks for your work!

May I ask more details on the def _mask function?

What is the functionality of this function?

Thanks for help

Could you please provide the checkpoint?

Hi @chamorajg, thanks for the uploading pytorch implementation of the 'Unsupervised Monocular Depth Learning in Dynamic Scenes' paper. Were you able to reach the numbers mentioned in the paper for Kitti/Ciyscapes?

Thanks for your work.
It seems tthat he activation function in MotionVectorNet is missing if I understand correctly. Looking forward to your reply.

Thank you for this work,

Could you share the script to generate Kitti and Cityscape training datasets?

Hello! In this line, shouldn't 'motion_features' contain the rgb and predicted depth of second frame as well? So index 0 would be 1 in the mentioned line, correct? Please feel free to correct me in case I've understood it incorrectly

Hi!
Thank you for your generous sharing！

I can't get an acceptable result on KITTI, neither depth or motion.
Have you ever obtained results comparable to those in the paper？

Hi @chamorajg, thank's for your pytorch version of original ones, may i ask that did you can reproduce the results in KITTI or waymo dataset?