NMKD Stable Diffusion GUI vs Drag Your GAN

NMKD Stable Diffusion GUI runs text-to-image and image-to-image generation locally on your Windows PC using your own GPU. The itch.io download bundles dependencies so you skip manual Python setup, and version 1.11.0 ships with a Stable Diffusion 1.5 model in a 3.3 GB package or a 1.5 GB build if you bring your own checkpoints.

Unlike browser-based generators that queue jobs on remote servers, NMKD keeps everything on your hardware with no data collection. It adds LoRA training on 8 GB GPUs, InstructPix2Pix editing, RealESRGAN upscaling, and CodeFormer or GFPGAN face restoration in one desktop app. The author reports sub-2-second generations on an RTX 3090 and under one second on an RTX 4090.

Windows hobbyists, game texture artists, and creators who want uncensored local control are the main audience. Built-in prompt queues, tileable output for game assets, custom VAE and checkpoint loading, and a malware scan on downloaded models make it a full workstation tool rather than a quick online demo.

Drag Your GAN lets you reshape a generated image by dragging handle points to target spots on the canvas. Instead of training on labeled pose data or 3D priors, DragGAN moves pixels inside a pretrained GAN's image manifold so edits stay photorealistic even when occluded regions get hallucinated. The project page hosts the SIGGRAPH 2023 paper, demo clips, and links to open source code.

Most GAN editors rely on fixed training annotations or 3D meshes, which limits what you can deform. DragGAN pairs feature-based motion supervision with discriminative point tracking so handle points follow your cursor while staying on the object surface. That makes it useful for nudging animal poses, facial expressions, car angles, and landscape layouts without retraining the generator.

Researchers, graphics students, and computer vision engineers use it to prototype interactive image editing workflows. The Max Planck Institute page distributes the paper PDF and GitHub repo under a non-commercial Creative Commons license, with contact details for the lead author.