Welcome to the ARTSMARTAI roadmap!

Ability to promt for logo shape, decribe image and text.  Give multiple options to choose.  Once chosen a logo that is close, use additonal prompts to fine tune.  Once locked in, ability to prompt to change angle, lighting etc.

Can images be generated without backgrounds like png transparency? A lot of images are extremely detailed even utilizing an AI background removal tool may be challenging. Providing images with transparent backgrounds would be an awesome feature. 

2000x2000 ?

Create a mode trained to produce images like photography!

UPDATE:
This feature has been fully implemented.

https://docs.artsmart.ai/collection

This means that we can create a character, for example an animated drawing, and that same character can be represented in different scenarios and doing different activities. For example, I create a character who is a girl with a unicorn horn and I can make an illustration playing in the park, another with the same character brushing their teeth in the bathroom, etc.

ability to save settings, such as, negative prompts inputs, resolution slider

The Ai model has an issue with producing hands at the moment.  We are researching ways to resolve this! 

Outpainting is the process of taking an original image and using AI to expand the borders seamlessly with new genrated imagery.   

This is great for situations where you need specific dimensions and your source image is too small.  

Add more painting styles and artists to help craft prompts.

Ability to Create Full Length Portrait : currently it only restricts to showing portraits above torso level. no legs?

Have a toggle option to indicate that I want the output to be very close to the input image. 

UPDATE: Implemented! 

https://docs.artsmart.ai/collection

For futher feedback please join us in the discord channel.  We've made a channel specifically for further needs on this feature! 

https://discord.com/channels/1049147793408794674/1088522630530863164


Hi, 

I have multiple projects and I generate images for illustrate my novels. It will be great to have the possibility to create folders and subfolders. For example, LMS, Book A - Character A , Book B, Website A, Website B, etc.

On profile page or another page when we can manage our creations.

DUPLICATE: Batch/Multiple Image Generations

Let's say that you are pleased with the output that artsmart AI generated but what if you want to have more variety for that particular output? this feature is a must. The final output will generate 4 more variants for the user to select. Of course, users who wish to generate more variants will need to spend more on their credit.

Allow users to choose from a variety of custom models, for different styles.

Maybe a dropdown menu or a tag word to use certain group of words with the prompt.

Make tunes usuable in the playground so users have full suite of prompt tools to used on their Tune.

We are currently working on a way to store tunes.  

Each tune is a large nueral network that requires a lot space. 

Would be great to have an option to have a text to video generation.

Current face enhancer is mostly useless. Please look into creating something similiar to CodeFormer that actually gives good results with strenght sliders: https://huggingface.co/spaces/sczhou/CodeFormer

create a business website design logo

Creating a batch of images would help in reducing the time of creating single images and finding the ideal image. 

Often the AI produces more subjects then you request.  Even when requesting a single subject in the prompt.  

We are looking for solutions to give more control over this. 

Training and tagging data to tune image models will be the biggest effort of ai art and ai photography.  We'd like to create a marketplace to allow artists and tune engineers to monetize their efforts. 

The marketplace would allow our users to create extremly highend tunes (image models) and recieve payment each time the model is used for generation in artsmart.ai

Possibility to delete an image from the own galery.

Upload a bulk of images and upscalin it at once.