Feature Requests

When we tap the remove button and then click on background, the background should remain as same as png form and there should come two option one is keep and another is add background so that if the user wish to save as png form he can save it and download it or as well if the user wish to add background he/she can add background by clicking add background option

mobili di design che escono dalla testa di una donna in stile rinascimentale. Fondo pulito

急にエラーとなり生成されなくなった。有償に値しない。

Estaría genial poder tener varias opciones de formatos para descargar las imágenes que hemos modificado con Relleno generativo en la web. Ya que al descargarse automáticamente en .png se modifica el tratamiento del color.

It's not clear that submissions take generative credits. I didn't know it did until I was out of credits. Also, if submitting ideas takes credits, all of those versions should be accessible. I used all the credits before I knew I was and have nothing to show for it. It would be nice to have a history of submissions. I clicked on the submission tab and it was empty. 

Image generator is almost trash. Only creates 10% of the input - idea. You guys are too far behind that image generation game. 

I can't get a glowing sphere to work very well.  Here is my prompt:   Closeup of a young redheaded woman holding a brightly glowing purple sphere in her hands, shown in profile, dark room, showing purple glow on her face, no lights in room I tried adding the last 3 items, but they didn't help. Attached is a sample of the closest image that I got. 

I find it weirdly hypocritical to allow image generation of mermaids but not harpies, gorgons, or medusa. I havn't tried much else but I was hoping for some visuals for a dnd session. 

I would like to propose an idea to remove an 'On-screen text remover' feature. It is very difficult to flawlessly remove the on-screen texts from a picture using the 'Brush' tool. Therefore, I would like to request the team to include such a feature which could automatically detect all the onscreen texts from the picture and remove it leaving behind only the image.

I am finding some older favorite images less appealing now that i have a better hang of Firely. It seems that a button could be had directly on each favorite that one might like to delete would be a nice inclusion. 

I would like to see a way or perhps multiple ways to insure sharpness in all stages of image generation in Firefly of in the PSD version. To me, it is critical to have the sharpness and resolution intact no mater what alterations are made. If there could be a way to always increase or at least maintain a level of  sharpness, it would be amazing. I'm sure it can be accomplished. It just might be a goal to achive oveer time. I want it now. 

Ich würde mich sehr freuen, wenn ich zukünftig die Möglichkeit hätte, mit Text zu Bild Eingaben Fotos von Orten zu verändern. Wenn ich z.B. im Sommerurlaub eine faszinierende Aufnahme eines Bergsees gemacht habe, würde ich daraus auch gerne ein Winterbild mit viel Schnee machen. Außerdem wäre es schön, wenn ich genau diesen Ort mitten in der Nacht mit Vollmond, Milchstraße und Sternen zeigen könnte oder beim Sonnenuntergang, im mystischen Nebel, usw.. Manchmal mache ich auch im Winter schöne Aufnahmen von Sonnenuntergängen, bin aber enttäuscht, dass an diesen Tagen kein Schnee liegt. Das würde ich auch gerne mit KI ändern. Momentan dürfen ja keine KI-Bilder von konkreten Orten erstellt werden. Ändert sich das in Zukunft?

Ist es möglich, dass ich zu Fotos Objekte, Menschen oder Tiere hinzufüge? Wenn ich z.B. einen Geburtstagskuchen fotografiere, wäre es toll, wenn ich mit Text zu Bild Prompts noch einen Blumenstrauß, Geschenke, Kerzen usw. hinzufügen könnte. Außerdem wäre es großartig, wenn ich bei Naturaufnahmen oder Fotos von Menschen Tiere wie Rehe, Hunde, usw. hinzufügen könnte.

Just want to say that the new version (3) of Firefly is awsome. thank's for the update!!Martin

Give the generative ai a memory to give the user the ability to get various versions of a set scene then using future prompts to be able to alter the selected scene/format

There needs to be a one-time popup or something to direct individuals to the correct Firefly image generation page https://firefly.adobe.com/ before they post to these community forums. This will prevent prompt requests spam.

Treinar a geração de imagens em mais dimensões:O programa é incrível, mas o algoritmo não compreende a ideia de imagens invertidas (de ponta cabeça) juntas, por exemplo, entrelaçar um triangulo para cima e um para baixo.Se alguém sabe como fazer isso, por favor, compartilhe.Valeu ADOBE!  

There needs to be more offered to the free tier for (esp as most in this cat. are students and poorer creative-types) with limited resources 💻🎨. Who agrees?!..    Adobe staff, please take note 👀. How about hosting a better free tier by allowing users to opt-in to ads and get twice the amount of generative compute per month? 💡 There are so many talented creative-types out there with very little money to afford a package, but they can offer their usage data to you in return, which is also valuable for a company like Adobe. User data and user interactions collected are literally worth money 💰, let's not forget that.    So here's my suggestion: provide an ad-supported free tier with expanded resources 🙌 Make Abode better for its users, please!...  Who agrees with this idea?!

As a pixel art designer, I think it would be good to have a larger pixel art database since it has a lot of potential, with the styles I realize that it is quite good, so just with a little more training I could get optimal results that could be used as a visual reference to work or simply get good ideas

Cela serait hyper pratique de pouvoir générer des objets sur des fonds blancs ou transparents pour faciliter le détourage après.

Add a structural Reference to the Generative fill in Photoshop like in Firefly.

Access to my favourites from the text to image page would be very useful.

歴史認識が浅くプロンプトに書いた通りにならない。16世紀侍は日本人しかおらずまた一般的に男性を指す。しかしこの史実を打ち破り虚偽の画像を作成する。金銭対価見合わないツールである。

Create cake with happy anniversary message for Mamta and Kishore  

To have access to previously generated iterations in a personal library to keep track of credits used.