This paper uses three visual AI technologies to improve the quality of pictures.
1. Image super-resolution
Image super-resolution is to upgrade a low-resolution image to a high-resolution image. According to the resolution, it can be divided into 2 or 3 times of super-resolution. Image super-resolution is based on the pixels of the original image. Filling and expansion, so as to adjust the resolution without destroying the effect of the original image. Image super-resolution can be a good solution to viewing images under different resolution displays. Meet 2K, 4K display.
Example picture 
2 times super resolution
3 times overscore
2. Super-resolution enhancement
Super-resolution enhancement is mainly based on the above super-resolution to repair and enhance the image at the pixel level, improving the resolution of the image and improving the clarity of image details. It has a good repair effect for low-resolution and blurry pictures. Tried for all scenarios, regardless of image content.
Example picture 
2 times super-resolution enhancement 
3 times super-resolution enhancement
3. Face repair
Face repair is to repair the face image quality of face pictures, improve the clarity of face details, and optimize the quality of face images. For other Content images are less effective.
Original image 
after face restoration
The restoration of photo quality, based on the current visual AI method, can repair the pictures of various scenes and adapt to the current screen resolution. This article mainly introduces the quality of pictures, and the quality restoration based on pictures. Video is also a general restoration method. In this way, the old or low-resolution video can be well restored, and it can be played at a higher level under the current LCD screen.