You Won’t Believe Your Eyes: Combining Cloudinary’s Generative AI Transformations

Cloudinary’s AI features are more than powerful, they’re also user-friendly. Combining robust APIs and SDKs with generative AI provides developers with the easiest way to start editing images. This blog post offers some example use cases that demonstrate how these features can be progressively combined to create stunning new variations of original images.

Sometimes, companies have an existing library of images they want to reuse for new experiences. This process may involve removing text or an object and changing the image dimensions. Simply removing an object may sound easy enough for a Photoshop expert, but expanding the image is where things get tricky. 

Here’s an image of a house in the woods. We want to remove the car to highlight the property’s seclusion. We also need to reset the format from portrait to landscape for use in a banner:

Loading code examples

Let’s emphasize the solitude of the house:

Loading code examples

Loading code examples

Loading code examples

Our generative features’ most common use case is correcting and normalizing images supplied by vendors or subscribers as product or property listing images. Often, these businesses need to learn the condition of the photos they will receive. 

In this example from a travel and tourism site, the agency managing a hotel listings for their client has uploaded a beautiful but overly compressed and underexposed image of a couple on a tropical beach at night:

Loading code examples

Loading code examples

Loading code examples

Loading code examples

Many retailers need their product catalog to have a uniform look and feel for each product. Often, they receive products with different dimensions, quality, and backgrounds. In this example, we’ll adjust a product shot to have a common plain background and a drop shadow and crop it to a standard size.

Loading code examples

Loading code examples

Loading code examples

Loading code examples

Now, let’s look at how all our generative AI features work together to create a new image from your original. For this example, we’ll take an image of an “open” sign for a business and edit it to make an ad for Cloudinary AI:

Loading code examples Loading code examples Loading code examples Loading code examples Loading code examples Loading code examples

Of course, all of these images can be delivered in the best possible format and quality by using Cloudinary’s f_auto and q_auto. This ensures the user experiences the fastest possible loading time without sacrificing quality. Let’s look at the last example with f_auto and q_auto added:

Loading code examples

The resulting image on Chrome is a 78KB AVIF image, compared to the 250KB JPEG original. Quite a savings in size for two simple API commands.

When it comes to editing and creating new variations of your original images, the possibilities are limitless with Cloudinary’s AI transformations. These use cases span from the practical to the imaginative, demonstrating the versatility of our features. No matter what your image and video enhancements may be, you can be confident that Cloudinary transformations will integrate into your automated workflows, simplifying your processes, and accelerating your time to market, and freeing up teams to be more creative.

Sign up for Cloudinary today.