I’ve been experimenting with AI image generators like Midjourney, Stable Diffusion, and Leonardo AI to create realistic images of humans for social media and marketing. But I’ve hit a roadblock: generating consistent and accurate human figures across sessions is incredibly difficult. It’s not just me, right?
I’ve noticed that certain words or contexts trigger filters or lead to nonsensical results. For instance, I tried to generate an image related to sleep and found that the word ‘bed’ in my prompt seemed to throw things off completely, leading to bizarre or filtered outputs saying it’s explicit. I’ve also encountered inconsistencies in anatomy, with some features coming out distorted.
I understand the need for safety measures, but sometimes the restrictions feel too broad and limit creative exploration in non-harmful ways. It’s like these tools are rapidly evolving, but generating realistic depictions of humans in various situations still has a long way to go.
Have you run into similar issues or frustrating limitations when trying to generate images of people? What have your experiences been like with specific keywords or scenarios? Have you found any prompts or techniques that help overcome these issues? I’d love to hear your thoughts and see if this is a common experience!