Generative AI—Experimental generation of digital assets

Client

Internal project

Year

Ongoing

Services

Virtual photography
Conceptualisation & Visualisation
Augmented Photo Retouching

Collage of three images, at the top a Mazda MX5 car in grey on white background, on the bottom left an AI generated grey Mazda MX5 car on the side of the road with the drive taking a photograph of the Toscan landscape, on the bottom right an AI generated black and white image of the grey Mazda MX5 car on the road exiting a tunel with an exited driver at the wheel.

Car placed in Tokyo city landscape - AI generated

Car placed in city landscape - AI generated

Side by side, on the left the image of a real burgundy dress on white background and on the an right AI generated image of a young women wearing the burgundy dress as shot on a alee in Paris.

Side by side, on the left the image of a real burgundy dress on white background and on the right an AI generated studio portret of a young women wearing the burgundy dress.

Two side by side AI generated studio shot images of green and white Nike Air Force One High shoes, the left image on a green background and the right image on a neutral grey background.

Two side by side AI generated studio shot images of red and white Nike Air Force One High shoes, the left image on alight grey background and the right image on a cool medium grey background.

Collage of three images, top left an AI generated grape with wine splashes on white background, bottom left a studio shot AI generated of a wine bottle nex to a wine glass with a bottle pouring wine in the glass, on the right the image has been altered by adding a label with the wine splashed grape and the naming Golden Grape.

Generative AI will fundamentally change digital and e-commerce asset production along with all other aspects of content generation.

At Octagram, we have always strived to stay at the forefront of digital production technologies in order to meet the highest expectations coming from our clients, which are some of the most renowned brands in the world.

Since the inception of generative AI, we have continuously experimented with different tools and platforms to understand the current capabilities and potential applications in the e-commerce space, production of digital assets, and alteration of product shots.

The examples showcased above were generated in early 2023. Therefore, they should be taken with a grain of salt at the moment of reading this text. For sure, the generative AI models will have evolved extremely fast, and the images included on this page will most probably be outdated at the time of reading this page.

We believe that generative AI will have a huge impact on generating e-commerce photography as well as expand the creative capabilities of all designers.

In the gallery you’ll find examples of:

Images of existing products placed in an AI-generated context:
– The car image is an accurate 3D representation of the real vehicle and it’s placed in AI-generated landscapes (with or without a driver).
– The red dress is a real life product image placed on an AI-generated model and background.
Images entirely generated by AI:
– The Nike Air Jordan 1 shoes (early experiment controlling colours, shoe position, and background – controllability might improved significantly since oru experiment).
– A wine bottle and glass with an AI- altered label and logo design.
– A room setting with a couch.

Though these examples are not spectacular, our aim was not to generate the most realistic images, but to identify applicable ways of applying the power of generative AI to real-life product photography and the e-commerce space.

DISCLAIMER: We are aware that generative AI models have been trained on a vast amount of data, some of which might be copyrighted. The images displayed here are either real products or AI-generated. The products, shapes, and symbols depicted in these images resemble or may be indistinguishable from real products. The copyrights for these elements belong to their legal owners, and the images are intended for illustrative purposes only.

⸺ Layout localisation

Layout localisation refers to the process of translating the assets to various languages, using the translations provided by the client or produced by Octagram.

⸺ Motion graphics

Motion graphics or motion design, is the process of creating visual elements that convey motion or change. It uses a variety of animation techniques to bring static images and text to life.

⸺ Digital production

Digital production refers to the process of generating multiple assets based on a master creative/visual, by adapting it to different layout formats and/or localising/translating the copy for various countries & regions.