Thanks for your great work on the scene reconstruction from multi-view images. Currently, I want to follow your evaluation protocol and have some questions about the evaluation code.
Specifically for ScanNet data, I am wondering if I only need to specify the extracted mesh path (e.g., the path for the provided mesh in this repo) and run "scannet_eval/evaluate.py" file to calculate the metric.
Dear author,
Thanks for your great work on the scene reconstruction from multi-view images. Currently, I want to follow your evaluation protocol and have some questions about the evaluation code.
Specifically for ScanNet data, I am wondering if I only need to specify the extracted mesh path (e.g., the path for the provided mesh in this repo) and run "scannet_eval/evaluate.py" file to calculate the metric.
Bests, Runsong