search

lliuz / ARFlow

The official PyTorch implementation of the paper "Learning by Analogy: Reliable Supervision from Transformations for Unsupervised Optical Flow Estimation".
MIT License
249 stars 49 forks source link

occu_mask #41

Open lt573152873 opened 1 year ago

lt573152873 commented 1 year ago

during training, Occu_mask whether there is gradient return? when occu_mask is zeros, the loss is small;

if need occu_mask -> occu_mask.detach() ?