Integration for automatic KI-based image generation – presono's knowledgebase

Permissions

The following group permissions are required to use DallE:

With the “DallE image generation” permission, users in the group can generate images themselves. With the right “See all DallE requests” they can also see the requests of other users and save the images themselves in presono or download them.

Generate images

To generate a new image, click on the “New” button at the top right of the content structure and select “Create new image with Dall-E”.

Now you can generate a new image in the upper area (1) and see the history (2) with all previously generated images in the lower area.

Generate a new image

If you want to generate a new image, you have the following options to choose from:

Input/Prompt: Here you describe the image you want to generate. For example, “palm trees by the sea”. The more details you describe, the less random the generated image will be. You can also describe the style of your image. The input must not exceed 3800 characters.
Format: here you can choose between landscape and portrait format as well as a square image.
More information: here you will find a detailed description of all fields and functions.
Number of images: here you can determine how many images should be generated.
Advanced settings: you can make further settings here:

Quality: The quality of the generated image (not the resolution). HD creates images with finer detail and better consistency across the entire image.
Style: The style of the generated image. Vivid causes the model to create more hyper-realistic and dramatic images. Natural causes the model to generate more natural, less hyper-real looking images.
Refine input: Tells the AI whether the input should be refined or not. Due to the nature of AIs, disabling this option is not completely reliable as it is just a query that is added to the input.

Once you have made all your settings, you can click on “Generate image”.

A loading bar will appear and your image will be generated:

As soon as the results are available, they are added to the history and are available for one hour.

You have the following options and information for each image:

1. you will of course see your generated image.

2. here you can see the remaining time that your image will be available.

3. you will also see the exact and in this case refined input

4. here you can see the settings that you have transferred

5. here you can save the image directly in presono

6. or simply download it

7. click here to apply the parameters (all settings) for a new generation.

Of course, there may also be errors on the side of the AI, which you will then see directly in the response. In this case, you can simply repeat your request by clicking on “ Adopt parameters” and rephrasing your request, for example.

History

All generated images are available for one hour in the system. You can see the current status in the history at the bottom of each request:

You will see the title for each request, which is the original input. Also the status:

the image has been saved,
an error has occurred,
the image is no longer available (the time has expired)
or how long it is still available.

The overview on the right shows who sent the request and when it was sent. You can expand all requests to view the details. Simply click on the corresponding entry:

For images that have already expired, you can no longer see the image itself, but you can apply the settings for a new request by clicking on “Adopt parameters”.

For images that have been saved, you can see the image and also navigate directly to the image (if you have access to it) or download the image. Of course, you can also apply the settings for a new request by clicking on “Adopt parameters”.