How to use the Flair AI neural network to create interesting images of objects

What is Flair AI

Flair AI is a neural network that can be used to generate images for various tasks. The tool is useful for creating and designing advertising images, drawings for presentations and social networks, products for cards in online stores and other similar purposes.

Using this service, you can quickly create a series of pictures in the same style. This is useful for those who do not have the means to work with professional photographers and designers. Flair AI allows a person without special skills to create, for example, an image of a bottle of perfume against the backdrop of a beautiful landscape or studio in 10–15 minutes.

Just upload a photo of the item and enter a text query describing the desired result, after which the neural network will generate a picture with a background and effects. The finished file can be saved to your computer or mobile device.

Flair AI is currently only available in English, but with the help of hints in the settings you can figure it out even without extensive knowledge of the language. The resulting images can be used for commercial purposes.

What the Flair AI neural network can do

Flair AI visually determines the image of the central object and then creates the environment: adding the correct shadows, reflections and even smoke, if you specify such a feature. First, the system will automatically remove the background from the photo of the item you uploaded.

The neural network is able to adapt the lighting to the specified parameters and tries to generate the most natural result. But if desired, the product picture can be stylized.

Dozens of Templates are available in the settings menu, which can be edited to suit your purposes in the Editor section. The finished result can be supplemented with new elements. The item “Placement” contains frequently used options, but you can also enter your own text. For example, “on a round platform” (circular platform), “on a plate” (plate), “on a wooden table” (wooden table), “on the grass” (grass).

You can create an image without any ready-made templates at all. To do this, you can create a layout design yourself: write a new description of the background, add the necessary additional elements, specify their color, scale and location.

The Flair AI interface is easy to use and most users can figure it out. For many tasks, it is enough to know how to upload a photo of an object, where to fill in the “Placement”, “Surrounding”, “Background” and “Custom”.

The free version has all the basic tools available, with which you can generate up to 30 images. A subscription expands the range of available elements, but you can try most of the features without it.

How to use the Flair AI neural network

First you need to open the website Flair. A.I. and click on the Get Started button on the main page. The neural network is available in Russia without a VPN. A subscription will cost $10, but you can only pay for it with a foreign card.

The service will ask you to register an account - this can be done through email or social networks. Next, in the Assets section, upload an image of the item using the Upload Product Photo button. It is advisable to select a high-quality photo with an object on a uniform surface so that the neural network can successfully cut it out of the frame.

Then choose a ready-made background template or make your own. Add a description of the object and indicate its location relative to other objects in the picture. Next, describe the background and environment.

To correctly describe the scene, just select the above-mentioned Placement, Surrounding, Background, or write down your entire request. For example, “dropper bottle on a platform surrounded by smoke, in front of curtain” - “a bottle with a pipette on a platform surrounded by smoke, in front of the curtain.”

You can add geometric objects, flowers, drinks, sports equipment, accessories to the image, animals, decor. To do this, in the Elements section, select the desired object. If you need to post an image of a person, go to the Humans section. In the settings you can specify the size, location and color of the inserted elements.

After preparing the request, click the Generate button, after which the finished result will appear to the right of the frame assembly field. The Number of results setting allows you to select from 1 to 4 image variations.

All additional pictures will appear to the right of the very first one. To view, just scroll the slider at the bottom of the screen. If you don’t like the final versions, you can run the generation again to correctly assemble all the necessary elements in the frame.