Skip to main content

RadioGPT: The "world's first" radio station with artificial intelligence

The Scribble Diffusion neural network turns your sketches into pictures

The Scribble Diffusion neural network is a type of artificial intelligence (AI) that can take simple doodles and turn them into high—quality drawings. This neural network is based on a deep learning algorithm developed by Google researchers in 2019.
 
The algorithm works by first analyzing a set of high-quality images to learn about different artistic styles and techniques. He then uses this knowledge to transform a simple set of doodles into a drawing corresponding to the desired style. This process is carried out using a method called "image-to-image conversion", which involves converting an input image into an output image using a neural network.

How does Scribble Diffusion work? 
Scribble Diffusion uses a unique approach to image-to-image translation called "diffusion". In traditional image-to-image translation, the neural network directly maps the input image to the output image. However, this can lead to loss of detail and quality of the final image.

Propagation, on the other hand, breaks down the process of converting an image into an image into several stages. At each stage, the neural network adds more detail and complexity to the image. This process is repeated several times until the final image is obtained.

This approach allows the neural network to create high-quality images that exactly match the desired style. It also allows the user to adjust the amount of detail in the final image, giving them more control over the final result. 
 
How to Use Scribble Diffusion?
 ➤ Opening Scribble Diffusion;
➤ Draw whatever your heart desires;
➤ We write a description of the depicted object in the line below the drawing;
➤ Click "GO" and wonder what kind of artists we are.
"Cat"
 
 "Adventure Time"
 

"Beautiful woman"

In a word, experiment.

Popular posts from this blog

Microsoft Launches Bing NEW with Built-in ChatGPT

C hatGPT and Bing are a new era of search engine s "This is a new day for search," said Microsoft CEO Satya Nadella. Microsoft has been trying to get you to use Bing for 13 years now, but you don't want to, so its global market share is still in the low single digits. Now the company is doing everything possible to better compete with Google. At a press conference, Microsoft announced the integration of the Open air GMT-4 model into Bing, which has long been rumored, providing a ChatGPT -like job in the search engine. The company is also launching a new version of its Edge browser with these new artificial intelligence features built into the sidebar. Bing ChatGPT includes the most up-to-date data As expected, the new Bing now has the option to start a chat on its toolbar, which then takes you to a ChatGPT-like dialog. One important point should be noted here: while the OpenAI ChatGPT bot was trained on data that only covers 2021, the Bing version is much more up-to-date...
Read more

Midjourney

Generating images using the Midjourney neural network Our friends from the Simple Figma channel have prepared for you a simple instruction on how to use the Midjourney neural network.
Read more

Calligrapher — neural network for generating handwritten text

T he possibilities of artificial intelligence are growing at an incredible pace, while it is becoming more and more accessible. An example of this is ChatGPT , which has become a real problem for many educational institutions whose students have successfully delegated their academic responsibilities to him. But technology does not stand still. One of the leading experts in the field of machine learning, Sean Vasquez, developed the Calligrapher service.ai, capable of generating handwritten text based on printed input. The service has nine different styles available, which allows you to use sliders to adjust the output text by speed, stroke width and legibility. Calligrapher.ai Like a human hand, it outputs each letter based on the IAM On-Line Handwriting Database, which contains more than 1,700 samples of human handwriting. The service contains 13049 lines of text and can generate up to 86272 handwriting samples with a vocabulary of 11059 words. Unlike handwriting-mimicking fonts Calli...
Read more