FST
FST copied to clipboard
usr922
[NeurIPS'22] Learning from Future: A Novel Self-Training Framework for Semantic Segmentation.
Awesome work, but a little issue: Exploring a wider future, can it be regarded as an approximate equivalent form of a large batch size?
Missing pretrained/resnet101_v1c-e67eebb6.pth When run the ./train.sh. I want to ask where to download the resnet101_v1c-e67eebb6.pth
Hi, authors. Thanks for your great work! After reading the paper, the equation (1) confuses me. Since in the original Mean Teacher framework, the update equation is written as $\phi_{t+1}...
