Image generation models produce identical faces because training objectives focus on realism and prompt matching but do not explicitly encourage identity diversity — adding diversity as an optimization objective via reinforcement learning solves this.

6:34our own models also what was interesting that uh this is not an image quality problem the problem that I mentioned before like we are asking the model to generate faces and and certain number of faces and it keeps generating same faces identical almost identical faces over and over again. So quality wise, image quality wise, if you look at the pixels and noise and everything, it looks realistic. But the m missing piece was

7:02that those models, base models, amazing models had not really learned to create truly distinct identities because uh existing training objectives focus heavily on the realism and matching the user prompt, but they don't explicitly encourage diversity. uh between people and that is very important. So that observation led us to uh a simple question what if the identity or facial appearance or any diversity you know kind of itself

7:38becomes an optimization objective. So that led into the idea behind disco. Instead of you know creating a completely new T2I model, we kept the underlying model and fine-tune it with reinforcement learning. Um we designed rewards that encourage several things simultaneously. For instance, different people within an image should have distinct identities. You don't want to, you know, kind of duplicate create duplicate faces. That's something we call in the paper as intraim image