Winning Solution in NTIRE19 Challenges on Video Restoration and Enhancement (CVPR19 Workshops) - Video Restoration with Enhanced Deformable Convolutional Networks. EDVR has been merged into BasicSR and this repo is a mirror of BasicSR.
So the wiki says to prepare the datasets via lmdb.
I prepared my custom datasets using the create_lmbd.py from BasicSR.
The question is how do I incorporate them into test_Vid4_REDS4_with_GT.py because I can't find the lmdb being imported and my custom datasets cause my GPU to crash.
In testing, it reads images instead of lmdb files.
LMDB files are used during the training for faster IO speed. You can also use image folders, but the IO may be the bottleneck.
So the wiki says to prepare the datasets via lmdb. I prepared my custom datasets using the create_lmbd.py from BasicSR. The question is how do I incorporate them into test_Vid4_REDS4_with_GT.py because I can't find the lmdb being imported and my custom datasets cause my GPU to crash.
Thanks