Detailed demonstration pytorch framework implementations old photo repair (GPU)

I. Environmental description

Environment	Version
System	windows10
graphics card	RTX3060
CUDA	11.1.1
cuDNN	8.1.1
Python	3.8.8
torch	1.9.0+cu111
torchvision	0.10.0+cu111
torchaudio	0.9.0

cuda installation reference: tensorflow-gpu2.4.1 detailed steps of installation and configuration

Algorithm model download address: link: https://pan.baidu.com/s/1uGNXd5g2dKjb4zgckxiLdA
Extraction code: vqv6

Second, install dependent modules

1.Create a new virtual environment.

2. Install torch, torchvision, torchaudio modules.

pip3 install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio===0.9.0 -f https://download.pytorch.org/whl/torch_stable.html

3. Install dependent libraries.

pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple

4. Store the photos to be repaired in the ./Bringing-Old-Photos-Back-to-Life/input folder

and execute the following command:

python run.py --input_folder C:\Users\Administrator\Desktop\Bringing-Old-Photos-Back-to-Life\input --output_folder C:\Users\Administrator\Desktop\Bringing-Old-Photos-Back-to-Life\output --GPU 0

Execution result:

Generate repaired photo path:

Effect comparison:

photo before restoration

photo after repair