In recent times, a lot of innovations are happening in the field of artificial intelligence, particularly in the generative AI category. Since Stable Diffusion, the text-to-image art generator, positioned itself as a pure open source, a slew of interesting models have been released at regular intervals by individual researchers.
In this article, you will learn about one such model that turns your simple or ugly sketch into an awesome image.
Diffuse the Rest
Diffuse the Rest is a prompt-guided image-to-image model that turns your simple sketches into awesome high-quality realistic art.
Instead of the sketches, you can also upload photos from the Internet and turn them into high-quality AI art using the prompt.
Steps to Turn Sketches into Awesome Images: Image-to-Image AI
- Go to Hugging Face
- Do your Sketch or Upload an Image
- Give Prompt
- Click “diffuse the f rest”
- Download the AI-Generated Images
1. Go to Hugging Face
Diffuse the Rest is hosted in Hugging Face spaces. To use it, you need to go to Hugging Face.
Hugging Face is an open-source community like GitHub for pre-trained deep learning models, mainly aimed at natural language processing (NLP).
2. Do your Sketch or Upload an Image
Once you are in the Diffuse the Rest Hugging Face space, you can start your sketch in the given white box.
You will be given a brush, eraser, and colors to choose from. You can also increase or decrease the size of your brush.
You can do very simple sketches or you can go with a detailed sketch.
Instead of a sketch, you can also upload any of the photos that resemble your idea.
3. Give Prompt
After doing the sketch, you need to enter a prompt corresponding to the sketch.
Since this is an image-to-image process, a simple prompt can give you better results.
If it was not an image-to-image model, you need to craft your prompt well according to prompt engineering.
4. Click “diffuse the f rest”
After entering the prompt, you need to simply click the “diffuse the f rest” button.
Now, the diffusion process will take place. On the interface, you will only be able to see dark matters. But, in the background, the diffusion algorithm does its best to produce the requested image.
5. Download the AI-Generated Images
After 20 seconds of processing, you will get four realistic images pertaining to your sketch.
Now, you can use the play/ pause button to navigate through the 4 AI-generated images and use the download button to download the image you want.
Not safe for work (NSFW) warning is a notable thing while working with AI art generators, particularly Stable Diffusion (SD).
The image that you are going to generate using Diffuse the Rest might pose a risk of reinforcing or exacerbating societal biases, pornography, and violence.
This is because Stable Diffusion was trained on the LAION-5B dataset, which is a collection of non-curated image-text-pairs scraped from the Internet.
AI art generators never fail to cater to users’ needs. Diffuse the Rest is not an exception.
Every designer or illustrator can quickly turn their basic idea into a high-quality realistic art piece in less than a minute.
1. Can I use the image generated with Diffuse the Rest?
Diffuse the Rest model is released under the CreativeML Open RAIL-M license. According to this license, the authors claim no rights to the output image.
Hence, you can freely use the images generated by the Diffuse the Rest model. But, you must not violate the license policies, such as causing harm to a person, spreading misinformation, or targeting any groups. You can read the full policies here.