Diffusion based Generative Models

Stable Diffusion

• The dynamic team of Robin Rombach (Stability AI) and Patrick Esser (Runway ML) from the CompVis Group at LMU Munich headed by Prof. Dr. Björn Ommer, led the original Stable Diffusion V1 release. They built on their prior work of the lab with Latent Diffusion Models and got critical support from LAION and Eleuther AI.

• Stability-AI Stable Diffusion
  • Stabiity-AI Company Site
• Runway-ML Stable Diffusion
  • Runway-ML Company Site
• CompVis Stable Diffusion
  • CompVis Lab
• LoRA Models

Processing Tools

• T2I-Adapter
  • Article: T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models
  • Lightweight T2I-Adapters to align internal knowledge in T2I models with external control signals, while freezing the original large T2I models. In this way, we can train var- ious adapters according to different conditions, achieving rich control and editing effects in the color and structure of the generation results. Further, the proposed T2I-Adapters have attractive properties of practical value, such as com- posability and generalization ability. Extensive experiments demonstrate that our T2I-Adapter has promising generation quality and a wide range of applications..
• Automatic1111 Web-UI
  • Instalation
• Web-UI ControlNet v1.1
  • ControlNet v1.0
  • ControlNet v1.1
  • [Article: Adding Conditional Control to Text-to-Image Diffusion Models] (https://arxiv.org/pdf/2302.05543.pdf)
• Comfy-UI
  • Modular stable diffusion GUI backend. UI for designing a workflow/pipeline.
• chaiNNer
  • GUI Image processing pipeline.

Depth Tools

Pose/Skeleton Tools

• OpenPose
• HAN: An Efficient Hierarchical Self-Attention Network for Skeleton-Based Gesture Recognition

Upscaling / Super-resolution

• Real-ESRGAN
  • Article: ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks
  • Article: Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data

Finetunning LoRAs and Special Checkpoints

• DreamBooth: Fine Tunning Text-to-Image Diffusion Models for Subject-Driven Generation

• LoRA: Low-Rank Adaptation of Large Language Models
  • Article: LoRA
• Kohya’s GUI

  • Gui for Training LoRAs

  • Article: LoRA

• Civit-AI
  • Collection of LoRAs for StableDiffusion note that some are designed for specific checkpoints.