release full fusion results for tanks and temples
Fusion results for all scans using default parameters in eval.py
(except that indoor scenes have --min_geo_consistent=3
).
Each point cloud contains 100M~300M points.
Due to the large size of some scenes, I compressed them. Please decompress before visualization.
Put under results/tanks/points
.
Use python visualize_ply.py --dataset_name tanks --scan {scan name}
to visualize.