1. Open Image

Open a Photoshop document and select an image

2. Open Neural Filters

Go up to the top and select Filter > Neural Filters

3. Select Super Zoom

The Neural Filters panel will load on the right side of Photoshop. Under the Beta Filters window toggle the Super Zoom filter on.

4. Place Zoom

Adjust Zoom within the window as desired

5. Adjust Options

Adjust other options as needed. In this instance I turned on Enhance Image Details and Enhance Face Details

6. Output

Select Output Method

Zoom Image

Zoom as it describes allows you to determine both the magnification and crop. Using the + icon we can increase the magnification by factors of 1. As this happens the bounding box about with the source image changes. Dragging this bounding box determines the overall crop of the image.

Enhance Image Details

A checkbox option which enables the AI to start inputting details that would otherwise be lost.

Remove JPEG Artifacts

The next checkbox is to Remove JPEG Artifacting that becomes more noticeable on a zoom. JPEGs are naturally compressed down and depending on the artifacting, a zoom will become more and more difficult. Enabling this option when working with JPEGs is usually a good idea, however always check to make sure you’re not losing any quality in the details. This has its own tool in the Neural Filters list however can be seen to have limited success depending on the images and settings chosen.

Noise Reduction

Noise Reduction is just like it sounds. Removing noise from an image can be key especially when zooming in, so Adobe added in a noise reduction option to help compensate for this.

Sharpen

Another issue that occurs in zooming in is focus becomes more and more important. In order to help alleviate any issues in focus (albeit focus should be perfect in camera but things happen and maybe the magnification goes beyond the setup), Adobe has included a sharpening tool to help mitigate the issues.

Enhance Face Details

Another checkbox option is to Enhance Face Details tells the Sensei AI to enhance the face of anyone it detects. I assume that this is due to the AI having to run different processes for facial details over general details.