KITTI.md

Preparing KITTI Dataset

1. Download Dataset

To obtain the KITTI dataset:

1. Visit the KITTI Raw Data official website.
2. Register for an account to access the download links.
3. Choose the specific dates and drives you need.
4. Download the following components:
   • Synchronized and unrectified data
   • Calibration files
   • Tracklets
5. After downloading, organize the files according to the file structure shown below.

File Structure of Raw Data

Project_path/Kitti/raw/
├── 2011_09_26
│   ├── 2011_09_26_drive_0001_sync
│   │   ├── image_00
│   │   ├── image_01
│   │   ├── image_02
│   │   ├── image_03
│   │   ├── oxts
│   │   ├── velodyne_points
│   │   └── tracklet_labels.xml
│   ├── 2011_09_26_drive_0002_sync
│   │   └── ... (similar structure as 0001_sync)
│   ├── ...
│   ├── calib_cam_to_cam.txt
│   ├── calib_imu_to_velo.txt
│   └── calib_velo_to_cam.txt
├── 2011_09_28
│   ├── 2011_09_28_drive_0001_sync
│   │   └── ... (similar structure as 0001_sync)
│   ├── ...
│   ├── calib_cam_to_cam.txt
│   ├── calib_imu_to_velo.txt
│   └── calib_velo_to_cam.txt
└── ...

2. Install the Development Toolkit

pip install pykitti

3. Process Raw Data

To process the raw KITTI data, use the following command:

# export PYTHONPATH=\path\to\project
python datasets/preprocess.py \
    --data_root data/kitti/raw \
    --dataset kitti \
    --split 2011_09_26 \
    --split_file data/kitti_example_scenes.txt \
    --target_dir data/kitti/processed \
    --workers 32 \
    --process_keys images lidar pose calib dynamic_masks objects

The extracted data will be stored in the data/kitti/processed directory.

4. Extract Masks

To generate:

• sky masks (required)
• fine dynamic masks (optional)

Follow these steps:

Install SegFormer (Skip if already installed)

⚠️ SegFormer relies on mmcv-full=1.2.7, which relies on pytorch=1.8 (pytorch<1.9). Hence, a separate conda env is required.

#-- Set conda env
conda create -n segformer python=3.8
conda activate segformer
pip install torch==1.8.1+cu111 torchvision==0.9.1+cu111 torchaudio==0.8.1 -f https://download.pytorch.org/whl/torch_stable.html

#-- Install mmcv-full
pip install timm==0.3.2 pylint debugpy opencv-python-headless attrs ipython tqdm imageio scikit-image omegaconf
pip install mmcv-full==1.2.7 --no-cache-dir

#-- Clone and install segformer
git clone https://github.com/NVlabs/SegFormer
cd SegFormer
pip install .

Download the pretrained model segformer.b5.1024x1024.city.160k.pth from the google_drive / one_drive links in https://github.com/NVlabs/SegFormer#evaluation .

Remember the location where you download into, and pass it to the script in the next step with --checkpoint.

Run Mask Extraction Script

conda activate segformer
segformer_path=/path/to/segformer

python datasets/tools/extract_masks.py \
    --data_root data/kitti/processed \
    --segformer_path=$segformer_path \
    --checkpoint=$segformer_path/pretrained/segformer.b5.1024x1024.city.160k.pth \
    --split_file data/kitti_example_scenes.txt \
    --process_dynamic_mask

Replace /path/to/segformer with the actual path to your Segformer installation.

Note: The --process_dynamic_mask flag is included to process fine dynamic masks along with sky masks.

This process will extract the required masks from your processed data.

5. Human Body Pose Processing

Prerequisites

To utilize the SMPL-Gaussian to model pedestrians, please first download the SMPL models.

1. Download SMPL v1.1 (SMPL_python_v.1.1.0.zip) from the SMPL official website
2. Move SMPL_python_v.1.1.0/smpl/models/basicmodel_neutral_lbs_10_207_0_v1.1.0.pkl to PROJECT_ROOT/smpl_models/SMPL_NEUTRAL.pkl

SMPL-Nodes (SMPL-Gaussian Representation) requires Human Body Pose Sequences of pedestrians. We've developed a human body pose processing pipeline for in-the-wild driving data to generate this information. There are two ways to obtain these data:

Option 1: Download Preprocessed Human Pose Data

We have uploaded preprocessed human pose data for a subset of KITTI scenes to Google Drive. You can download and unzip these files without installing any additional environment.

# https://drive.google.com/file/d/1eAMNi5NFMU8T7tjQBT_jzxeX-yJRwVKM/view?usp=drive_link
# filename: kitti_preprocess_humanpose.zip
cd data
gdown 1eAMNi5NFMU8T7tjQBT_jzxeX-yJRwVKM

unzip kitti_preprocess_humanpose.zip
rm kitti_preprocess_humanpose.zip

Option 2: Run the Extraction Pipeline

To process human body poses for other KITTI scenes or to run the processing pipeline yourself, follow the instructions in our Human Pose Processing Guide.

6. Data Structure

After completing all preprocessing steps, the project files should be organized according to the following structure:

ProjectPath/data/
  └── kitti/
    ├── raw/
    │    ├── 2011_09_26/
    │    │   ├── 2011_09_26_drive_0001_sync/
    │    │   │   ├── image_00/
    │    │   │   ├── image_01/
    │    │   │   ├── image_02/
    │    │   │   ├── image_03/
    │    │   │   ├── oxts/
    │    │   │   ├── velodyne_points/
    │    │   │   └── tracklet_labels.xml
    │    │   ├── ...
    │    │   ├── calib_cam_to_cam.txt
    │    │   ├── calib_imu_to_velo.txt
    │    │   └── calib_velo_to_cam.txt
    │    └── ...
    └── processed/
         ├── 2011_09_26_drive_0001_sync/
         │  ├──images/             # Images: {timestep:03d}_{cam_id}.jpg
         │  ├──lidar/              # LiDAR data: {timestep:03d}.bin
         │  ├──ego_pose/           # Ego vehicle poses: {timestep:03d}.txt
         │  ├──extrinsics/         # Camera extrinsics: {cam_id}.txt
         │  ├──intrinsics/         # Camera intrinsics: {cam_id}.txt
         │  ├──sky_masks/          # Sky masks: {timestep:03d}_{cam_id}.png
         │  ├──dynamic_masks/      # Dynamic masks: {timestep:03d}_{cam_id}.png
         │  ├──fine_dynamic_masks/ # (Optional) Fine dynamic masks: {timestep:03d}_{cam_id}.png
         │  ├──objects/            # Object information
         │  └──humanpose/          # Preprocessed human body pose: smpl.pkl
         ├── 2011_09_26_drive_0002_sync/
         └── ...