Artificial intelligence (AI) and machine learning are major topics in e-commerce these days. We want to show you four features of the Feed Image Editor app that we use machine learning for. There is a lot of work and time behind their release, but you can use them with one click. So which functions do we talk about?
Upscale
The small size of the images can stop the product from being advertised on comparison engines. They require quality images with minimum pixel dimensions, and failing them can be problematic for you. When you enlarge them in a graphics editor, the quality is lost, and they are scattered.
Did you know…?
The machine-based resolution enhancements found in the Feed Image Editor can increase the size and quality of the images. It works in bulk and only for those images where improvements are needed.
Background removal
One of the standard requirements of comparison engines is a white background without disturbing elements. Many products are photographed in the context of a specific environment. An example is a piece of furniture shown directly in the room. Such images are attractive, but you should use them only as alternative images. The main image of the product must have a white background.
If only a couple of them are problematic, you can remove them manually. But if there are hundreds or thousands of them, it's incredibly time-consuming. Remember that even bulk background editing tools don't always remove them perfectly. The product images are very colorful. They have complicated shapes, obscure edges of the object, or holes in them. In the Feed Image Editor, we try to take all these factors into account and improve background removal for images in the context of e-commerce.
Did you know…?
The terms machine learning and artificial intelligence are often confused or misused. Machine learning is an AI sub-area based on processing large amounts of data and searching for various dependencies, with which the machine learns (trains) to achieve a set goal. The undeniable advantage of machine learning is its use for solving challenging tasks. It achieves high accuracy, which often exceeds people's abilities.
Poor image quality detection
Even technically high-quality images that meet the requirements of comparison engines and get into advertising may not look good on the content page. Many factors, such as low contrast between the background color and product color, dispersion, blur, etc., are no longer controlled by them.
The Feed Image Editor helps detect a wide range of these problems and suggests how to fix them.
Watermark detection
Watermarks are one of the most common factors why product images are not allowed in advertising. Watermarks have a protective function in images, and in this form, online stores often receive them directly from suppliers. But comparison engines do not support them.
The Feed Image Editor helps you detect watermarks no matter how large they are, where they are located in the image, or whether they are outside the product or overlap.
TEST THESE FEATURES FOR FREE
See for yourself how to use the four features mentioned in the Feed Image Editor app. You get it free for the 30 days.
If you have any questions, contact us by e-mail or arrange an online meeting with product owner Marek Daňko.
Add new comment