imagine the AI can understand the nature of the object in the picture and therefore manipulates the object(s) accordingly. For instance, in the concept image of a pair of shoes put side by side, the user should be able to ask Photoshop to remove one of the shoes by selecting one and having Photoshop generate the rest of the image previously blocked by the show in front according to the prompt of the user or an optional auto-generated suggestion.
Furthermore, it would be wonderful if the user can proceed to ask Photoshop to change the angle of view. For instance, view from the top, right side, left side, or even bottom.
This would be really useful in prototyping, or in creating images to be merged for a 3D panoramic view from a 2D image.
I tried the delete and fill option but it came out bad. Picture us attached.
I also tried a reflective background and it actually generated a reasonably good reflect of the sole of the shoes. (image also attached) This shows that it is possible for generative fill to be used to change the angle of view of images in the photos and also generate replacements.
It would be more amazing if the user is then able to describe what the generated replacement should look like.
I should be paid for this kind of idea. No AI tool can so this yet.
AdChoices
.png){kind=link}
