The purpose of Adobe Express
Hello
I have tried Adobe Express a couple of times now and would like to talk about my experience.
So what exactly is the purpose of Adobe Express? I thought it is supposed to be an AI support to create images. If that is the case let me share my experience with you:
I was trying to create two images:
1. Image in anime style, cozy feeling, a couch with table on a carpet in the front. Cat under the table. Some pillows on the couch. Umbrella next to the couch.
From the images I one that seems OK but the umbrella flies like Mary Poppins above the couch. How can I fix this? There should be a way to mark elements that I do not like and where I can tell the AI to fix this specific issue. There is something that can create variations but it is not where you would look so I spent a substancial amount of time looking for the function. I gave up looking for it this time. Why is the tool so hidden?
Other images might be better but I do not like the colouring/art style. There should be a function where I can feedback the AI "this image is good but I do not like the style. give it a more cozy feeling" but that function is also missing.
2. Image of a small city map from above. Make several buildings stand out as post office, bank, cafe, university. streets with some traffic lights at intersections. Hand drawn art style.
This commans is even more disasterous. First the roads are all over the place. The white road lines are going all directions. Extremely messy and totally not realistic.
I understand this is very challenging for AI. I uploaded a hand drawing of what I want. My idea was that the AI will recognise the roads and buildings on the map and transforms it into a digital art. It does not. I do not really understand why I can upload images at all? When I want to remove the background I can just use Photoshop.
Also I wish there were a function where I can add elements to the picture and the AI is merging colour and style
In summary: Adobe Express is absolutely useless at this point for generating images. I tried many times to make it work but the interface has a terrible accessability, a lot of simple features are missing, and the AI part is often mediocre. AI from NVIDIA can create stunning images from paint drawings and even photoshop has a more powerful generative AI than Adobe Express. Shouldn't it be the same at least?
Maybe I am too dumb to use it properly but that would also indicate that the interface is not accessible enough. At this point I am trying to use AE every couple of months and quit again because it is basically useless. Then I usually end up doing things from scratch in Photoshop.
Unfortunately I lack the IT knowledge to provide more detailed suggestions but my advice would be: polish interface, make it easier to do "spot fixing". Enable recognition of images and make it so they can be changed in art style, light, colouring etc. This is to some extend already possible in PS so it should not be too difficult.
