Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
wuyy 21e5952197 | 1 year ago | |
---|---|---|
assets | 1 year ago | |
compressai | 1 year ago | |
compressai.egg-info | 1 year ago | |
dataset | 1 year ago | |
docker | 1 year ago | |
docs | 1 year ago | |
examples | 1 year ago | |
models | 1 year ago | |
tests | 1 year ago | |
third_party/ryg_rans | 1 year ago | |
CONTRIBUTING.md | 1 year ago | |
LICENSE | 1 year ago | |
MANIFEST.in | 1 year ago | |
Makefile | 1 year ago | |
NEWS.md | 1 year ago | |
Readme.md | 1 year ago | |
mypy.ini | 1 year ago | |
pyproject.toml | 1 year ago | |
requirements.txt | 1 year ago | |
setup.py | 1 year ago | |
test_video.py | 1 year ago | |
train_video.py | 1 year ago |
CompressAI_MindSpore
├── models
│ ├── dcvc.py
│ ├── ssf2020.py
│ └── STPM.py
├── dataset
│ ├── vimeo90k\
│ └── vimeo90k.py
├── test_video.py
└── train_video.py
conda create -n compressai_mindspore python==3.8
conda activate compressai_mindspore
pip install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html
pip install range-coder==1.1
pip install pytorch-msssim==0.2.1
pip install scipy
pip install opencv-python
pip install Pillow
pip install mindspore
cd /CompressAI_MindSpore/
pip install -e .
vimeo90k_dir
}python3 train_video.py -d {vimeo90k_dir}/vimeo_septuplet --epochs 100 --model ssf2020 -lr 0.0001 --lambda 1e-2 --batch-size 4 --save --dir_checkpoint {your_saved_dir}
python3 train_video.py -d {vimeo90k_dir}/vimeo_septuplet --epochs 100 --model dcvc -lr 0.0001 --lambda 1e-2 --batch-size 4 --save --dir_checkpoint {your_saved_dir}
python3 train_video.py -d {vimeo90k_dir}/vimeo_septuplet --epochs 100 --model ssf2020 -lr 0.0001 --lambda 1e-2 --batch-size 4 --load_checkpoint {your_saved_checkpoint} --save --dir_checkpoint {your_saved_dir}
python3 train_video.py -d {vimeo90k_dir}/vimeo_septuplet --epochs 100 --model dcvc -lr 0.0001 --lambda 1e-2 --batch-size 4 --load_checkpoint {your_saved_checkpoint} --save --dir_checkpoint {your_saved_dir}
vimeo90k_dir
}python3 test_video.py -d {vimeo90k_dir}/vimeo_septuplet --model ssf2020 --lambda 1e-2 --test-batch-size 1 --load_checkpoint {your_saved_checkpoint}
python3 test_video.py -d {vimeo90k_dir}/vimeo_septuplet --model dcvc --lambda 1e-2 --test-batch-size 1 --load_checkpoint {your_saved_checkpoint}
link:https://pan.baidu.com/s/1cGNl3zh8cuj9cX5n7wMIdQ
password:1234
loading dataset from splitfile: /code/CompressAI_MindSpore/dataset/vimeo90k/sep_testlist.txt
Loading saved_checkpoint/CompressAI_MindSpore_ssf2020/checkpoint_best_loss.ckpt
inference: 0 Loss: 1.954 | MSE loss: 0.001 | PSNR: 31.606 | Bpp loss: 1.51
inference: 1 Loss: 2.277 | MSE loss: 0.001 | PSNR: 30.244 | Bpp loss: 1.66
inference: 2 Loss: 1.446 | MSE loss: 0.000 | PSNR: 34.515 | Bpp loss: 1.22
inference: 3 Loss: 2.013 | MSE loss: 0.001 | PSNR: 30.992 | Bpp loss: 1.50
...
...
inference: 54 Loss: 1.219 | MSE loss: 0.000 | PSNR: 38.413 | Bpp loss: 1.13
inference: 55 Loss: 3.964 | MSE loss: 0.003 | PSNR: 25.608 | Bpp loss: 2.18
inference: 56 Loss: 1.541 | MSE loss: 0.000 | PSNR: 34.541 | Bpp loss: 1.31
inference: 57 Loss: 1.087 | MSE loss: 0.000 | PSNR: 38.201 | Bpp loss: 0.99
inference: Average losses: Loss: 2.505 | MSE loss: 0.001 | PSNR: 32.126 | Bpp loss: 1.89
@article{li2021deep,
title={Deep contextual video compression},
author={Li, Jiahao and Li, Bin and Lu, Yan},
journal={Advances in Neural Information Processing Systems},
volume={34},
pages={18114--18125},
year={2021}
}
@inproceedings{agustsson2020scale,
title={Scale-space flow for end-to-end optimized video compression},
author={Agustsson, Eirikur and Minnen, David and Johnston, Nick and Balle, Johannes and Hwang, Sung Jin and Toderici, George},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={8503--8512},
year={2020}
}
@inproceedings{ivr,
title={Interpolation variable rate image compression},
author={Sun, Zhenhong and Tan, Zhiyu and Sun, Xiuyu and Zhang, Fangyi and Qian, Yichen and Li, Dongyang and Li, Hao},
booktitle={Proceedings of the 29th ACM International Conference on Multimedia},
pages={5574--5582},
year={2021}
}
The implementation is based on CompressAI, STPM, DCVC and ssf2020.
Yuyang Wu
No Description
Python C++ Text Makefile other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》