Segment Anything in Light Fields for Real-Time Applications via Constrained Prompting
Project Page: roboticimaging.org/Projects/LFSAM/
Authors: Nikolai Goncharov, Donald G. Dansereau;
Published at: Proceedings of the Winter Conference on Applications of Computer Vision (WACV) Workshops, 2025, pp. 1490-1496
Paper link
Create virtual environment and run the requirements:
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
Install SAM 2:
bash build.sh
Download model checkpoints
cd sam2_checkpoints
bash download_ckpts.sh
cd ..
Download UrbanLF dataset (6.5 GB free space, around 5 minutes waiting):
bash download_urbanlf.sh
python experiments.py ours_config.yamlfor our method. The result tensors and metrics will be put into./experiments/ourspython experiments.py baseline_config.yamlfor baseline method. The result tensors and metrics will be put into./experiments/baseline