zju3dv / NeuralRecon

Code for "NeuralRecon: Real-Time Coherent 3D Reconstruction from Monocular Video", CVPR 2021 oral
https://zju3dv.github.io/neuralrecon/
Apache License 2.0
2.03k stars 294 forks source link

The configuration used for your pretrained weights #92

Open jcliu0428 opened 2 years ago

jcliu0428 commented 2 years ago

Hi thanks for sharing your great work!

May I ask a question that what are the configurations you used for the pretrained weights you provided? e.g., the batch size and training epochs, the learning rate and the training split of ScanNet? Since I follow your default setting to train on ScanNet v2 train split but the performance(fscore) has some discrepancy of your pretrained weights, so I wonder whether this is due to different training configuration or split. Looking forward to your reply!