Artificial Intelligence is much more than what some users think when they hear this term and it is not exactly related to what we have been sold traditionally some Hollywood movies. The uses of artificial intelligence range from translating into other languages, to video games, including photography, recognizing objects, finding malicious software, among others. But, in addition, it can also be used to create images from words.
Bearing in mind that one of the usual uses of this technology is to recognize objects, it is not difficult to find platforms that, using Artificial Intelligence, allow us to create copyright-free and completely unique images. If you want to know some of the best platforms to create images through words, we invite you to try the following options.
DALL-E is an artificial intelligence program, whose name comes from Wall-E and the Spanish painter Salvador Dalí, that we can use to create images from descriptions, the more complete they are, the better results we will obtain. It uses 12,000 million parameters to interpret the text in natural language and creates realistic images with great detail. It is one of the platforms that offers us the best results, it is not free and is available through its website.
platform google machine learning that allows us to generate images based on a text is Google Brain. If we take into account that the Google Photos platform uses a neural network system capable of recognizing objects and animals, and that its development began more than a decade ago, although with other purposes than creating images from text. Google has limited the use of this platform to academic institutions, so the only way to access it is through our school, institute or university center from its website.
DreamStudio is a platform completely free that we can use to create images based on text. If we want the best results, it is recommended to use English instead of Spanish. We can download all the images that are generated through this website completely free of charge, although the resolution leaves a lot to be desired (512×512 pixels). Within the configuration options, we can modify the width and height of the image, the number of images to generate based on the text entered, among other aspects. We can access DreamStudio and create any image based on a text through the following link.
Another interesting option that we have at our disposal for free is Stable Diffusion Demo. Through this website, we can generate unique and unrepeatable images based on the text we enter. Unlike DreamStudio, where the result is generated in just a few seconds, with this website we have to wait a few minutes to obtain the images we are looking for. That yes, the results are more satisfactory than DreamStudio and do not give the sensation of being images with pasted cuts. Stable Diffsion Demo is available at this link.
Stable Diffusion UI
If we want to obtain the best results to generate images from texts, the best free option is Stable Diffusion UI, an application that we can download and install completely free for both Windows and Linux. If we want to get the most out of it, it is recommended to have a graphics card with at least 6 GB of memory, although without a dedicated graphics card, the process will be much slower, however, it really makes up for the wait, since the results are much better than the rest of the options that we have mentioned. We can download this application from its website on GitHub by clicking on the following link.
The application occupies only 220 MB, but, once we run the installer, we are going to need 20 GB of spacesince the installation will download a large number of files to be able to carry out this process locally, taking full advantage of our team, which will allow us to obtain the best results.