Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference

arxiv:2310.04378

Original source · Paper JSON

Self-supervised learning and vision · 2023-10-06

Cites (0) · Cited by (1)

Complete cleaned Markdown · Source Markdown including bibliography

Read in parts

1. Part 1
2. Part 2
3. Part 3
4. Part 4

Contents

• Latent Consistency Models: Synthesizing High-Resolution Images with Few-step Inference
• Abstract
• 1 Introduction
• 2 Related Work
• 3 Preliminaries
• 4 Latent Consistency Models
• 4.1 Consistency Distillation in the Latent Space
• 4.2 One-Stage Guided Distillation by solving augmented PF-ODE
• 4.3 Accelerating Distillation with Skipping Time Steps
• 4.4 Latent Consistency Fine-tuning for customized dataset
• 5 Experiment
• 5.1 Text-to-Image Generation
• 5.2 Ablation Study
• 5.3 Downstream Consistency Fine-tuning Results
• 6 Conclusion
• Appendix A More Details on Diffusion and Consistency Models
• A.1 Diffusion Models
• A.2 More Details on Consistency Models in ( Song et al., 2023 )
• Appendix B Multistep Latent Consistency Sampling
• Appendix C Algorithm Details of Latent Consistency Fine-tuning
• Appendix D Different ways to parameterize the consistency function
• Appendix E Formulas of Other ODE Solvers
• Appendix F Training Details of Latent Consistency Distillation
• Appendix G Reproduction Details of Guided-Distill
• Appendix H More few-step Inference Results
• Instructions for reporting errors