deepdriving

AI and ML techniques applied to the KITTI automotive dataset! This is mostly demonstrations of different computer vision and deep learning techniques for Visual-SLAM, 3D Reconstruction, and Pointclouds.

As described in the MonoReq repo...

Run the code below to move data_depth_annotated.zip files into the sequence directories:

MonoRec$ python data_loader/scripts/preprocess_kitti_extract_annotated_depth.py \
--output ~/KITTI/dataset/ \
--input ~/KITTI/data_depth_annotated.zip \
--depth_folder image_depth_annotated

KITTI Odometry Dataset is expected in the user home of your filesystem:

~/KITTI/dataset/poses/
               /sequences/00/image_0/
                            /image_1/
                            /image_2/
                            /image_3/
                            /image_depth_annotated/
                            /velodyne/
                         /01/
                         /02/
                         /...
                         /21/