DASR super resolution model that responds to various image degradations
This is the original image 128*171
If you stretch it to make it easier to see, the pixels will appear jagged.
When this is scaled 4 times (512*684) with DASR
If you compare the necklace and the background, I think you can see that it is quite clear.
There are all kinds of images in the world, and there are many kinds of deterioration. I’m trying to respond to that.
How to use
install
git clone https://github.com/The-Learning-And-Vision-Atelier-LAVA/DASR.git
cd DASR
execution
!python quick_test.py --img_dir='your_dir/your_image.jpg' --scale='4' --resume=600 --blur_type='aniso_gaussian'
Pretrained models are included in the repository.
The execution time is fast at 0.2 seconds (T4) for 256*256.
🐣
I’m a freelance engineer.
I write various articles about AI, so please take a look at my profile if you’d like.
If you have any of the following requests, please feel free to contact us.
I want to develop an AI service, I want to incorporate AI into my business to improve efficiency, I want to develop a smartphone application using AI, I want to
create an application using AR, I want to create a smartphone application, but I don’t know where to turn…
All of these can be ordered at reasonable prices, eliminating intermediate costs.
For business inquiries, please contact
I am creating applications using machine learning and AR technology.
I send machine learning / AR related information.