I experienced some difficulty keeping consistency in my images for this assignment. Sometimes regenerations of photos would leave out certain details that I had already clarified in the configuration for the GPT (i.e., color palette, art style, etc.). I also had to fix my character’s appearance multiple times and ended up not using some really awesome photos because they weren’t consistent enough with my already existing photos and ideas. In order to create the character in particular, I went through a similar process to when I created the concept for this world in the first place, asking my GPT for details that it would need to get a clear picture of her. Once I answered those questions, I was also able to envision her as a character and what her appearance and personality would be like, which made it easier to tailor my world’s story to fit into her life. Overall, this was easily the most fun I’ve had so far with the photo generation, and I’m excited to do this again as a starting point for the video assignment.
https://docs.google.com/presentation/d/1KAxAs679Akbx2DzYIgXLUOcGj7AiX98jlq5LeZxq4-k/edit?usp=sharing