ControlNet
An architectural extension for diffusion models that dramatically improves spatial consistency by conditioning image generation on strict structural hints.
Think of It Like This
Like giving an imaginative painter a strict architectural blueprint to trace over, ensuring the final fantasy castle actually has walls that make sense.
Text prompts are notoriously bad at describing exact poses or specific edge boundaries. ControlNet solves this by locking the original diffusion model weights and adding a small, trainable copy of the encoding layers. This allows users to pass in depth maps, canny edges, or human pose skeletons to perfectly dictate the geometric structure of the final output.