H3 Sex Doll Concept LoRA

A concept LoRA for Minimax H3, designed for text-to-video (T2V) and image-to-video (I2V) generation. In image-to-video workflows, it can preserve the original person's facial identity while introducing a realistic sex-doll appearance to the body when undressed. This creates a distinctive contrast between a familiar face and a synthetic doll-like body. Reference-to-video (R2V) compatibility has not yet been tested.

Experiment with different prompts, scenes, movements, outfits, and camera angles to explore the concept's capabilities.