Stability AI近日发布了其最新文本到图像生成模型Stable Diffusion 3的研究论文，该模型在排版和提示遵守方面表现出色，超越了当前业界领先的同类技术。Stable Diffusion 3采用多模态扩散Transformer（MMDiT）架构，实现了图像和语言表示的显著提升。相较前版本，SD 3在文本理解和拼写能力上有了显著的进步。这一研究成果标志着AI模型在创意生成领域的一大飞跃。
英文标题：Stability AI Releases Research Paper on Stable Diffusion 3
英文关键词：AI Model, Text to Image Generation, Stable Diffusion 3
英文新闻内容：
Stability AI has recently released a research paper on its latest text-to-image generation model, Stable Diffusion 3. The model demonstrates superior performance in layout and prompt compliance, surpassing current state-of-the-art technologies such as DALL·E 3, Midjourney v6, and Ideogram v1. Stable Diffusion 3 utilizes a multi-modal diffusion Transformer (MMDiT) architecture that significantly enhances the representation of both images and language. Compared to previous versions, SD 3 has made significant progress in text understanding and spelling capabilities, marking a major leap forward in the field of AI-generated creativity.
【来源】https://stability.ai/news/stable-diffusion-3-research-paper
Views: 0