Make folders to store raw data and processed data
mkdir nuscenes processed
cd processed
mkdir maptr maptrv2 stream_new stream
Download nuScenes V1.0 full dataset data and CAN bus expansion data HERE. Unzip them into the nuscenes directory
Download CAN bus expansion
# download 'can_bus.zip'
cd nuscenes
unzip can_bus.zip
Folder Structure
MapBEVPrediction
├── nuscenes/
│ ├── can_bus/
│ ├── maps/
│ ├── samples/
│ ├── sweeps/
│ ├── v1.0-mini/
│ ├── v1.0-test/
| ├── v1.0-trainval/
├── processed/
│ ├── maptr/
│ ├── maptrv2/
│ ├── stream/
│ ├── stream_new/
Prepare Pre-trained Model
cd MapTR_modified
mkdir ckpts
cd ckpts
wget https://download.pytorch.org/models/resnet50-19c8e357.pth
wget https://download.pytorch.org/models/resnet18-f37072fd.pth
Prepare nuScenes data
python tools/create_data.py nuscenes --root-path ../nuscenes --out-dir ../processed/maptr --extra-tag nuscenes --version v1.0 --canbus ../nuscenes
Using the above code will generate nuscenes_infos_temporal_{train,val}.pkl.
Folder structure
MapBEVPrediction
├── nuscenes/
├── processed/
│ ├── maptr/
| | ├── nuscenes_infos_temporal_train.pkl
| | ├── nuscenes_infos_temporal_val.pkl
│ ├── maptrv2/
│ ├── stream/
│ ├── stream_new/
Prepare Pre-trained Model
cd MapTRv2_modified
mkdir ckpts
cd ckpts
wget https://download.pytorch.org/models/resnet50-19c8e357.pth
wget https://download.pytorch.org/models/resnet18-f37072fd.pth
Prepare nuScenes Data
python tools/maptrv2/custom_nusc_map_converter.py --root-path ../nuscenes --out-dir ../processed/maptrv2 --extra-tag nuscenes --version v1.0 --canbus ../nuscenes
Using the above code will generate nuscenes_map_infos_temporal_{train,val}.pkl, which contain local vectorized map annotations.
Folder structure
MapBEVPrediction
├── nuscenes/
├── processed/
│ ├── maptr/
│ ├── maptrv2/
| | ├── nuscenes_map_infos_temporal_train.pkl
| | ├── nuscenes_map_infos_temporal_val.pkl
│ ├── stream/
│ ├── stream_new/
Prepare nuScenes Data
cd StreamMapNet_modified
python tools/nuscenes_converter.py --data-root ../nuscenes --newsplit --dest_path ../processed/stream_new
python tools/nuscenes_converter.py --data-root ../nuscenes --dest_path ../processed/stream
Using the above code will generate nuscenes_map_infos_{train,val}_{newsplit,}.pkl, which contain local vectorized map annotations.
Folder structure
MapBEVPrediction
├── nuscenes/
├── processed/
│ ├── maptr/
│ ├── maptrv2/
│ ├── stream/
| | ├── nuscenes_infos_temporal_train.pkl
| | ├── nuscenes_infos_temporal_val.pkl
│ ├── stream_new/
| | ├── nuscenes_infos_temporal_train_newsplit.pkl
| | ├── nuscenes_infos_temporal_val_newsplit.pkl