Image specialty models

These are not standalone models but rather conditioningConditioningAny extra input that steers generation beyond the prompt, such as a depth map, pose, edge map, or reference image. techniques that can be added to existing diffusion models (typically Stable Diffusion or FLUX) to give them stronger control over identity and style transfer. InstantIDInstantIDA technique for preserving a specific person's identity across generations from a single reference photo, without training a LoRA. is specifically focused on preserving a person's face from a reference photo while generating them in new contexts. IP-AdapterIP-Adapter (Image Prompt Adapter)A method for conditioning generation on a reference image, so a subject or style from a picture carries into new outputs. is a more general conditioning mechanism that lets you use any image as a style or content reference. Both are available as ComfyUIComfyUIThe dominant node-based interface for running diffusion models locally. Started by 'comfyanonymous' in January 2023, now stewarded by Comfy Org. custom nodes and are widely used in production workflows.

ControlNetControlNetA conditioning technique that lets you provide structural guidance (depth, edges, pose) to a diffusion model to control spatial layout., originally released in 2023, is one of the most important add-ons in the entire generative imagery ecosystem. It allows you to give a diffusion modelDiffusion modelA generative model that learns to produce images by training a neural network to reverse a noise-adding process. Generates by starting from random noise and iteratively denoising to produce a clean image. additional structural conditioningConditioningAny extra input that steers generation beyond the prompt, such as a depth map, pose, edge map, or reference image., a depth mapDepth mapA grayscale image encoding how far each pixel is from the camera, used as conditioning to control 3D structure., an edge map, a pose skeleton, a segmentation mask, to control the spatial layout of the output. Want to generate a person in a specific pose? Provide a stick figure. Want a specific architectural layout? Provide a depth map. ControlNet works with most Stable Diffusion variants and has been adapted for FLUX and other newer models. It is the technique that made AI image generation usable for tasks that require precise composition control.