Leveraging Midjourney’s generative AI, I developed a collection of photorealistic images that highlight women and athleisure fashion in athletic environments.
The art direction focused on close-up shots to accentuate fashion details, while keeping facial visibility to a minimum.
This self-directed project aimed to enhance my prompting skills and, if successful, compile a selection of these images for upload to Adobe Stock. The goal was to evaluate the requirements and limitations of AI-generated content on stock image platforms.
Designing photorealistic images with Midjourney offered a chance to both explore and challenge the acceptance of AI-generated visuals on platforms like Adobe Stock. By meticulously refining my prompts, I aimed to produce highly detailed and lifelike images while testing how these refined prompts met the specific standards and criteria of stock photo platforms.
This approach allowed me to better understand the requirements and limitations of photo banks and assess the potential of AI-generated content in the stock photo market.
Through this exploration, I identified several effective strategies for optimising AI-generated images for stock photo platforms. A key insight was to avoid rendering faces, which simplified compliance with complex model consent forms required by these platforms. Instead, I focused on creating close-ups that highlight the model’s attire and body movements.
This approach not only adhered to platform requirements but also provided a more compelling presentation of athleisure fashion. By emphasising clothing details and dynamic poses, I enhanced both the visual appeal and relevance of the images within the athleisure category.
Freelance designer in Hong Kong, specialising in graphic and web design, focusing on crafting impactful visual identities and developing websites for brands to connect with their audiences ➞
