Finding the right balance between accuracy and creativity is one of the most challenging aspects of training Generative AI neural networks. You may have experienced this yourself, with higher ChatGPT temperature settings that result in AI "hallucinations", where ChatGPT will invent sources or make factually incorrect statements. When it comes to generating product images, maintaining precision and realism is crucial, but so to is infusing artistic touches of imaginative lighting and scene compositions to capture attention and drive conversion. At Ecomtent, we've addressed this challenges head on, with our new product feature that allows customers to seamlessly toggle between two distinct AI modes: Precise and Creative.

‍

Precise Mode: In the precise mode, customers can expect their generated product images to be represented entirely accurately without 'hallucinations' common in AI generated images. To achieve this, our generative AI model in the precise mode is trained using advanced computer vision techniques and deep learning algorithms. It carefully analyses the uploaded image provided by the customer, learning to understand and capture the intricate details of the product. The model acquires the ability to reproduce the product's shape, colour, texture, and other relevant features with a high level of precision.

Customers would typically opt for the precise mode for uncommon or unique products, which otherwise Generative AI may add additional elements to given the presence of these elements in the majority of the data it was trained on (e.g. adding a handle to a handless mug).

‍

Creative Mode: In the creative mode, customers can expect a touch of artistic interpretation and variation while still maintaining a realistic visual presentation of their products. This mode allows for the generation of images with imaginative lighting, compositions or backgrounds, enhancing the overall aesthetic appeal of the output. Training the AI model for the creative mode involves a combination of computer vision and generative modelling techniques. The model is exposed to diverse datasets containing both product images and highly converting ecommerce content references. Through this training process, the model learns to incorporate additional elements such as lighting conditions, compositions, and contextual backgrounds into the generated images.

The creative mode is the preferred choice for customers seeking a touch of artistic flair and enhanced visual appeal in their product images while still maintaining a realistic presentation. This mode is particularly useful when customers want to create visually captivating images that stand out, whether for marketing purposes or to showcase their products in a more artistic and aesthetically pleasing manner.

Having these two models now seamlessly integrated into our user experience empowers the customer to make the right choice for their individual product and desired use case.  This feature just the latest of many which makes Ecomtent an industry-leading AI tool for ecommerce.

‍

See a youtube demo here: Ecomtent new features - YouTube

‍