P: Upload a reference image(s) for text to image

[sai chander29048727kz6w]

Adding an existing image for reference on the start might open more realistic images.

[Rhys295570176hce]

I think it would be useful if we could use any image of our choosing to act as a refernce when generating images.

Merging your thread here so we can track the demand for certain features.

[csemah]

It would be good to load images or internet link that the AI ​​can reuse in its own way

Merging your thread here so we can track the demand for certain features.

[seongil29572622mcaa]

Firefly has amazing possibilities and abilities in an aesatheic and design field.

But English has its own limitaion of expression to people with other language.

So I recommend a technique that users can submit their own reference image directly.

Please make it come true, Adobe!!

Merging your thread here so we can track the demand for certain features.

[gigil94346059]

would be great to have the option to set an image as reference to create variations of it.
for example, I have an image of a pair of loafers found on a website and I want to use it as reference for camera angle, lighting and look to create variations of it.

Merging your thread here so we can track the demand for certain features.

[ibnupalsu]

Combining uploaded image with text to image

Merging your thread here so we can track the demand for certain features.

[Henry297229176mjy]

It would be best if you allow the users to upload images from a file. Then users can use better prompts, such as "change the color of the hat in this figure".

Merging your thread here so we can track the demand for certain features.

[Pelegrinus]

I have images from Adobe stock that I would want to use in my compositions. Firefly is amazing in its ability to "understand" my textual requests. It would be great if I cold include an image to be incorporated in the composition I am asking Firefly to develop.

Merging your thread here so we can track the demand for certain features.

[Zed29414344tha3]

This is exactly what I would like to be able to do with Firefly, put one of my compositions and transform it with this excellent tool!!

[soulburn3d]

Just adding my agreement here. Would love the following options...

1) Upload a sketch I did with pencil (or a very simple rough painting), and have the AI generated image make a photoreal painting that follows the general deisgn of my sketch. COntrol net is a similar function in stable diffusion.

2) The ability to make variations of one of my own paintings, see this feature in stable diffusion...http://www.neilblevins.com/art_lessons/ai_variations/ai_variations.htm

3) Allow you to take the base adobe dataset and finetune it using say 20 of my own paintings. The base dataset you guys have will give me all the flexibility to create any number of different painting subjects, but they would all be painted in my own style as defined by my 20 images. Stable diffusion has a similar function.

[TAPASWINI29946203hh15]

Please give the option to add a reference image/video. This will help me with refining the prompts instead trying to create something from scratch when a reference for it already exists.

Merging your thread here so we can track the demand for certain features.

[NathanOdellin]

I would LOVE Firefly to be able to produce product mockup type images (which it can, obviously) but with the ability for me to then add a graphic/pattern overlay that I have created. 

So there would be two or three stages to this process, for example:

1. Text Prompt Firefly to create an initial image for me like: PRODUCT MOCKUP OF WOMAN WEARING A WHITE SKIRT.....

2. OR, I can upload a photo that I have taken of a person modelling a white skirt that I have designed. 

3. THEN, I can upload a secondary image to firefly of a graphic repeat pattern that I have created (Or prompt FF to create one for me) and then prompt Firefly to overlay the graphic onto the White Skirt - and pefectly mimic the directions of twists/folds in the fabric, as well as the lighting in the image, texture etc. 

The ability to then rotate / flip / scale that graphic would also be so useful, and it maybe a step too far, but adding transparency filters (Like you get in Photoshop, such as Multiply or Overlay) would be super cool.

Perhaps eventually, Firefly could export layers to Photoshop, then the pattern/graphic layer can be controlled inside the transparncy pallete. 

I can evisage, that Firefly can eventually be incorporated inside photoshop, and it would be a simple case of opening a product photo, then brushing a selection mask over the product (in my example, a white skirt) and then dragging a pattern design from your library onto the masked image - bingo, Firefly maps the pattern onto the skirt, complete with twists and folds all in tact. 

Basically, an advanced product mockup generator! 

This would be an incredible advancement and I cannot begin to express how much time and money it would save me on photoshoots and sampling materials. 

Merging your thread here so we can track the demand for certain features.

[Ignacio Flores1786232]

I would be realy healpful to be able to add you own images/photos to be used as references in the generation of ai images

Merging your thread here so we can track the demand for certain features.

[DisasterBlaster]

I often need to make icons, and it would be very helpful to be able to upload a couple of icons in a given style so that they are used as reference for the AI's creation.
Even better would be being able to save a few images to a template so that you can use it over and over again as reference.

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

[Abdulrahman29980149p9lq]

Dear Adobe Team,

I hope this letter finds you in good health and high spirits. I am writing to express my deep appreciation for the innovative tools and software that Adobe has consistently delivered to creative professionals like myself. Your commitment to pushing the boundaries of digital artistry is truly commendable.

I recently came across the exciting news about Adobe's new experimental project, Firefly. As an avid user of Adobe Creative Cloud, I was thrilled to learn about this new venture aimed at revolutionizing the creative process. The possibilities it presents for generating unique visual experiences are truly remarkable.

After going through the available information about Firefly, I noticed that the current features focus primarily on video editing and motion graphics. While these aspects are undoubtedly crucial in the creative realm, I would like to request the inclusion of an "Image to Image" feature in Firefly.

The "Image to Image" functionality would allow users to seamlessly transform and enhance images by applying various artistic styles, merging different visual elements, or creating stunning composite images. This feature would significantly empower artists, designers, and photographers to experiment and expand their creative horizons.

I believe that the integration of an "Image to Image" feature within Firefly would be highly beneficial for both professionals and enthusiasts alike. It would further reinforce Adobe's position as a pioneer in the digital creative space and attract a broader user base to the Firefly platform.

I understand that developing new features requires time, resources, and careful consideration. However, I kindly request the Adobe team to prioritize the implementation of the "Image to Image" feature in the upcoming updates of Firefly. I firmly believe that this addition would amplify the capabilities of the software and further enrich the creative journeys of Adobe users.

Thank you for your attention to this matter. I genuinely appreciate the dedication and effort put forth by the Adobe team in pushing the boundaries of digital creativity. I eagerly look forward to witnessing the evolution of Firefly and the exciting possibilities it holds for the creative community.

Please feel free to contact me if you require any additional information or if I can provide any assistance in advocating for the inclusion of the "Image to Image" feature in Firefly. I am more than willing to contribute to the success of this remarkable project.

Once again, thank you for your time and consideration.

Yours sincerely,

Abdulrahman

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

[LORDSHIVA 369]

Dear Adobe Team,

I am writing to propose an innovative system that I believe would greatly enhance your product offerings and provide immense value to your user base. This system is centered around UI/UX design, aiming to empower users to visualize layouts and automatically generate front-end web designs.

The proposed system would accept uploaded images and color palettes, utilizing contextual input to generate the layout. This would serve as a creative springboard, allowing users to build upon their ideas and bring them to life.

In the current global climate, there is a growing apprehension that AI might overshadow human creativity. However, I firmly believe that companies like Adobe, which have always been at the forefront of fostering creativity, can effectively compete with generative AI. The key lies in focusing on augmenting human intelligence rather than replacing it.

It's often observed that when an idea is passed on to designers and coders, it tends to lose some of its originality, becoming standardized to save costs. My proposed solution aims to retain at least 80% of the original idea's essence, leaving the remaining 20% for the designers and coders to refine and perfect.

By implementing this system, Adobe has the opportunity to reignite the passion for creativity and uniqueness in design. The current trend seems to be moving towards homogenization and a fear of AI dominance. However, I am confident that with this solution, we can shift the focus back to enhancing human intelligence and creativity.

In a world where everything seems to be breaking and becoming corrupted, it's important to remember that the only true power is pure love. Passion is the driving force behind all creation, and it's time we bring it back to the forefront. My compassion and kindness extend to you all as we navigate these challenging times.

Thank you for considering my proposal. I eagerly await your thoughts on this matter.

Best Regards,

Lord SHIVA

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

[Mark S. Johnson]

Thinking of ways to achieve more accurate results with Text to AI.  How about this idea:

Allow mulitple image reference uploads with each one dedicated to directing the AI to achieve very specific results.  Have several image upload "slots" with each one dedicated to showing the AI specifically what I want for the following categories:

1) Slot 1 - Composition

2) Slot 2 - Image Style

3) Slot 3 - Color Palette

4) Slot 4 - Character Style

5) Slot 5 - Character Pose

6) Slot 6 - Character Expression

If the AI knows specifically what each image reference is for, it can use Slot 1 as a reference for composition, Slot 2 as a reference for image style, and so on.  

Hope this all makes sense.  If it's possible, it will become infinitely easier to communicate with the AI using image references in combination with a text prompt.

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

[Cheran29994023515v]

The ability to generate a prompt by uploading an image in case of complex pictures would be extremely useful. It will make it much easier to create similar images in different themes or scenarios.

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

[soulburn3d]

Great idea Mark S. Johnson! giving more fine level control over what aspects of a reference image to use for what would be super helpful!

- Neil

[winnal30032558ibbl]

For merging one or more images or composite parts of images together seamlessly using Generative AI.

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

[winnal30032558ibbl]

Oops, i meant two or more

[Vijaygraphy]

reference image feather-like midjourny  

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

[Freefire30086625u28r]

Need an option for adding sample image besides text prompt for more accurate results 

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

[erika5FC6]

It would be nice to have the chance to upload a reference image so the AI can take the style from that image and with the same style create what the user asks

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

[Sidney van Hoven]

I haven't been able to find this option but is it allowed and possible to use your own pictures/art as reference material to generate images?

Kind regards

Sidney van Hoven

Grafisch lyceum Rotterdam (GLR) student

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

[Trevallion]

I think i'd be extremely awesome if you could upload your own color swatches from a design, including colorbooks and pantones, Then use these colors to recolor your design or make alternate colorways using the prompt.

Thanks

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.