Adobe could not stand aside against the backdrop of the popularity of DALL-E, Midjourney and other neural networks introducing your own service Firefly. This new AI-powered platform not only allows you to create images starting from a text description, but also to modify them immediately, changing various details. Let's go see how it works.
Adobe announces Firefly, a new generative AI model focused on creating images like Midjourney
Adobe Firefly is aextension of generative tools based on the artificial intelligence that Adobe introduced last year for Photoshop, Express and Lightroom. At the moment, the service is in beta and can be accessed for free by sending a request. Going forward, the company plans to integrate generative AI tools into its other applications. Adobe specifically claims that it did not infringe any copyrights to train its neural network. The company used non-copyrighted images licensed for educational purposes or from its Adobe Stock library.
Read also: How to use ChatGPT on Telegram: complete guide
In this way, the company protects itself from potential dissatisfaction on the part of the artists, since its neural network generates unique content and he will not copy any brand's product, as he has never seen it. At the same time, the company is ready to pay artists if they “feed” their content to its service. Indeed we know that neural networks train In the true sense of the word. Images created with Adobe Firefly will contain metadata indicating that they are partially or fully generated by artificial intelligence.
As for the possibilities of Adobe Firefly, they are very extensive. The neural network allows not only to generate images based on a text description and apply different styles to images existing, but also to perfect them. The examples presented show that it is possible to change clothes in public as well edit faces acting on emotions, the width of the smile, the opening of the eyes, the type of hair, etc. You can also constantly refine the requests of an image, adding more and more new effects, changing the background and so on.
