Wu Xiaodong

Results 1 comments of Wu Xiaodong

> Hi, > > This is the weights I trained on the ScanNet.v2 official train split with batch_size 10 for 120k iterations. (This did not use all the augmentations that...