How to edit only part of an image

Image-generating AI is very talented, it shows. However, it often happens that we are not satisfied with a particular item. Traditionally, this has not been easy to correct, if we ask,Amnesty InternationalAmnesty International Generator often creates a completely new image.

Dall.e 3 was updated in Spring 2024 which now allows you to edit only part of the image. Here we will see how to implement this generally unknown function.

Editing tools

Dall.e 3 can be accessed from the paid version of ChatGPTChatGPT. It is also possible to access it for free from Copilot from MicrosoftMicrosoft But at the time of writing this article, this version does not offer the editing tools we will discuss here.

The main point is that once you create any image with Dall.e 3, you will need to click on that image. From there, some tools emerge. The thing we're most interested in on the left, is the selection tool. When you click on it, you'll see a custom slider to affect the size of the selection brush.

Delete part of the image using Dall.e 3

Intelligently erasing part of an image usually requires the use of a specialized tool, as shown in this tutorial: How to Easily Erase Objects from an Image. Now Dall.e 3 can do it directly.

Test it on the image. Select the area you wish to delete. It appears covered in blue. Once the selection is made, write a prompt to the right, for example: “Delete this”, “Delete this part”… You can be more specific: “Remove cat”, “Delete tables from wallwall “…As you will see, the erasure is done while respecting the logic of the image.

Replace part of the image

Replacing part of an image follows the same logic: you have to select the area(s) to be replaced and then give the appropriate instructions: “Replace the helmet with a hat”, “Add a necklace to this woman’s neck”… It does not always carry out the required instructions exactly and you sometimes have to redo Formulate your request until you get the desired result. The fact remains that here we have a very useful tool because it overcomes the drawback that generative AI traditionally has and should lead you to produce images that are closer to what you specifically want to obtain.