Allow Multiple Input Images Per Character for Higher Likeness Accuracy

Currently, each Character supports only a single input image, and all image generations are based on that one reference. The problem is that one image only captures a single angle, expression, and lighting condition — so the model has limited information to work with, and appearance consistency drifts across generations.

Requested change:

• Allow multiple input images per Character instead of just one

• Use the full set as combined reference so the model can understand the character's appearance from multiple angles, expressions, and lighting

• Optionally let users mark a "primary" image while the rest reinforce overall likeness

Why it matters:

• A single image can't convey a full, consistent appearance — multiple references dramatically improve face and feature accuracy across different poses and scenes

• This is a natural extension of the existing single-input-image system, not a new pipeline

• It directly improves the consistency people are already asking for with persistent Characters and ongoing roleplay

Even a modest increase from one image to a small set would make a major difference in likeness accuracy. Thanks for considering it!