谷歌近日宣布，其开发的文本到图像生成AI模型Imagen 3正式向美国用户开放。用户可以通过AI Test Kitchen平台体验这一最新AI模型。据谷歌官方介绍，Imagen 3相较于前代产品Imagen 2在细节清晰度、光照丰富度和人工痕迹减少方面都有显著提升。
Imagen 3的开放标志着谷歌在AI图像生成领域的一次重大进展。该模型被认为在理解文本提示并将其准确转化为图像方面表现出色，且生成的图像更具创造性和细节。在谷歌的人工和自动评估中，Imagen 3的表现优于其他竞争模型，如DALL-E 3、Midjourney v6、Stable Diffusion 3和Stable Diffusion XL 1.0。
这一进展可能会对图像生成领域产生深远影响，并可能推动相关技术的进一步发展。随着AI技术的不断进步，未来我们可能会看到更多基于文本描述生成高质量图像的应用场景。
英语如下：
Title: “Google’s New AI Text-to-Image Model Imagen 3 Surpasses DALL-E, Available to U.S. Users”
Keywords: Google, AI-generated images, open
News Content: Google has announced the official opening of its text-to-image AI model Imagen 3 to U.S. users. Users can experience this latest AI model through the AI Test Kitchen platform. According to Google’s official introduction, Imagen 3, compared to its predecessor Imagen 2, significantly improves in terms of detail clarity, lighting richness, and the reduction of artificial traces.
The opening of Imagen 3 marks a significant advancement for Google in the field of AI image generation. The model is considered to perform well in understanding text prompts and accurately converting them into images, with generated images exhibiting greater creativity and detail. In both manual and automated assessments by Google, Imagen 3 outperforms competing models such as DALL-E 3, Midjourney v6, Stable Diffusion 3, and Stable Diffusion XL 1.0.
This advancement may have a profound impact on the image generation field and could further propel the development of related technologies. As AI technology continues to advance, we may see more applications that generate high-quality images based on text descriptions in the future.
【来源】https://www.ithome.com/0/788/864.htm
Views: 1