- Concept-Graphs
- Concept-Fusion (Todo)
- Open-Fusion (Todo)
- OpenMask3D
- OVSG (Todo)
- OVIR-3D (Todo)
- Semantic Abstraction (Todo)
- PLA (Todo)
git clone https://github.com/be2rlab/OpenSemanticMapping.git
cd OpenSemanticMapping
make build-concept-graphs
bash ./datasets/scripts/download_replica.sh
download from https://disk.yandex.ru/d/lAWeTD3SSNncaA
make prepare-terminal-for-visualization
make run-concept-graphs
Inside the docker you can edit the export files in export folder, and then run one of them
bash /export/export_concept_graphs_light_ram.sh
make build-openmask3d
sudo python3 ./datasets/scripts/transformations/transform_to_openmask3d_format.py /path/to/original/scene/folder /path/to/output/folder
make run-openmask3d
Inside the docker you can edit the export files in export folder, and then run one of them
bash /export/export_openmask3d_replica.sh
cd /opt/src/openmask3d/saved/experiment/visualizations/room1_mesh; python -m http.server 6008
Change "room1_mesh" with your dataset's name. Then open in browser: http://0.0.0.0:6008.
make build-openscene
python3 ./datasets/scripts/transformations/transform_to_openscene_format.py \
/path/to/original/scene/folder \
/path/to/output/folder
make run-openscene
Inside the docker run the command from /opt/src/
python ./scripts/feature_fusion/replica_openseg.py \
--data_dir /path/to/transformed/dataset \
--scene name_of_the_scene \
--output_dir /path/to/output/dir \
--openseg_model /path/to/openseg/model
Inside the docker you can edit the export files in export folder, and then run one of them
bash /export/export_openscene_replica.sh
Inside the docker you can run the following to visualize the results
cd ./demo/gaps
make
cd pkgs/RNNets
make
cd ../../apps/osview
make
/opt/src/demo/run_demo
To change the inputs you can open run_demo
In the opened window you can type queries!