GitHub - yiyunchen/CVQENet: NTIRE 2021 Challenge on Quality Enhancement of Compressed Video Track1

yiyunchen / CVQENet Public

Notifications You must be signed in to change notification settings
Fork 0
Star 7

NTIRE 2021 Challenge on Quality Enhancement of Compressed Video Track1

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
code		code
pretrainModel		pretrainModel
readme.txt		readme.txt

Repository files navigation

Test Code for CVQENet: CVQENet: Deformable Convolution-based Compressed Video Quality Enhancement Network, which took the 8th place in Track1 of the NTIRE 2021 Challenge on Quality Enhancement of Compressed Video.
More challenge details can be seen in https://github.com/RenYang-home/NTIRE21_VEnh

Environment:
pytorch:1.2

Prerequest: 
1. cd code/ops/dcn/
2. bash build.sh
3. python simple_check.py

Test:
1. cd code
2. option setting
   - pretrain: the path of the pretrained model, e.g. ../pretrainModel
   - nb1 : the number of reblock in FEM, our pretrained model is nb1 = 10
   - nb2 : the number of reblock in FQEM, our pretrained model is nb2 = 20
   - nf : the number of channel, our pretrained model is nf = 64
   - test_dir: the path of testing video images
   - image_out: the path to save the output image
3. run code, python test.py

Acknowledgement:
Our code is based on : https://github.com/RyanXingQL/STDF-PyTorch and https://github.com/xinntao/EDVR

Concat:
If you have any question, drop us a line at [email protected] or simply open a new issue.