Emu: Enhancing Image Generation Models Using Photogenic Needles in a Haystack

arxiv:2309.15807

Original source · Paper JSON

Self-supervised learning and vision · 2023-09-27

Cites (0) · Cited by (0)

Complete cleaned Markdown · Source Markdown including bibliography

Read in parts

1. Part 1
2. Part 2

Contents

• Emu: Enhancing Image Generation Models Using Photogenic Needles in a Haystack
• Abstract
• 1 Introduction
• 2 Related Work
• Text-to-Image Models.
• Fine-Tuning Text-to-Image Models.
• Fine-Tuning Language Models.
• 3 Approach
• 3.1 Latent Diffusion Architecture
• 3.2 Pre-training
• 3.3 High-Quality Alignment Data
• 3.4 Quality-Tuning
• 4 Experiments
• 4.1 Evaluation Setting
• Prompts.
• Metrics.
• 4.2 Results
• Effectiveness of Quality-Tuning.
• Visual Appeal in the Context of SoTA.
• Quality-Tuning Other Architectures.
• 4.3 Ablation Studies
• 5 Limitation
• 6 Conclusion
• 7 Acknowledgement