Latest Advances in Stable Diffusion Technology

4 Oct 2024

Explore the latest advancements in Stable Diffusion technology as presented by the Stability AI Applied Research team.

Additional Comparison between Single- and Two-Stage SDXL pipeline

4 Oct 2024

FID Assessment of Generative Text-Image Foundation Models

4 Oct 2024

Traditional FID scores may no longer effectively measure the quality of modern generative text-to-image models like SDXL.

Comparing SDXL and Midjourney v5.1 on PartiPrompts: Which AI Model Wins?

3 Oct 2024

A user study comparing SDXL and Midjourney v5.1 using the PartiPrompts benchmark reveals that SDXL slightly outperforms Midjourney in terms of prompt adherence

How Diffusion Models Generate Images

3 Oct 2024

Learn how Diffusion Models use Probability Flow ODEs, stochastic equations, and denoising score matching to generate images from noise.

Understanding the Limitations of SDXL in Image Generation

3 Oct 2024

The SDXL model faces limitations in rendering complex structures, achieving photorealism, addressing biases, handling concept bleeding, and improving long text

What’s Next for SDXL Development?

3 Oct 2024

This paper outlines potential SDXL improvements, including single-stage models, better text synthesis, transformer optimizations, and enhanced efficiency.

Improving Image Quality with Better Autoencoders

3 Oct 2024

Explore how the integration of an improved autoencoder and a multi-stage training process enhances the image quality in SDXL.

Multi-Aspect Training: Adapting SDXL for Real-World Image Diversity

3 Oct 2024

Discover how multi-aspect training in SDXL adapts to real-world image diversity by accommodating varying sizes and aspect ratios.