Code for evaluation.
guangkaixu opened this issue · comments
Hi, thanks for your great contribution to reconstruction. Is there any code for evaluation of predicted point cloud, like "C-l1", "IoU", "NC" and "F-score" mentioned in the paper?
Hi guangkaixu,
The evaluation code was adapted from Convolutional Occupancy Networks:
https://github.com/autonomousvision/convolutional_occupancy_networks/blob/f15b97da2c5995598537c8f832e52e95c0b09236/src/eval.py
You might also be interested in RetrievalFuse since they build on the same code:
https://github.com/nihalsid/retrieval-fuse/blob/main/util/mesh_metrics.py