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.
Your work is really fascinating and I've been using it extensively on my research project!
I see that you're building many custom operators such as the DCNv2 which have been copied from the mmcv repository.
Is there any reason why we should not use the mmcv code directly from their pip package? Specifically, I am using the EDVR model and I want to replace your DCNv2 function with the one from the mmcv pip package.
Hi @xinntao,
Your work is really fascinating and I've been using it extensively on my research project!
I see that you're building many custom operators such as the DCNv2 which have been copied from the mmcv repository.
Is there any reason why we should not use the mmcv code directly from their pip package? Specifically, I am using the EDVR model and I want to replace your DCNv2 function with the one from the mmcv pip package.