| 事情大家已经知道了：人工智能AI 图像使用“稳定扩散”人工智能AI 生成。人工智能的优势之一是单个模型可以生成多种图案，包括浮世绘、插图和照片。虽然有些图像非常精致且与人类绘图无法区分，但还是存在一些人工智能AI 输出的图像与众不同。这是因为人工智能从人类制作的大量图片和照片中学习图像，但人工智能AI无法学习常识、社会常识。 How 人工智能AI creates paint image（From NIKKEI Asia January 21, 2023 ） How does 人工智能AI generate images? Knowing how image generation and learning works, we may be able to find some tips on how 人工智能AI and humans can work together. Under the supervision of Makoto Shing, an applied scientist at 人工智能AI developer rinna Co,. Ltd. in Tokyo, the technology for creating images from words is laid out based on a paper on the image-generating 人工智能AI Stable Diffusion.人工智能如何生成图像？了解图像生成和学习的工作原理，我们或许能够找到一些关于人工智能和人类如何协同工作的技巧。在 人工智能AI 开发商 rinna Co. 的应用科学家 Makoto Shing 的监督下。Ltd. 在东京，根据一篇关于图像生成 人工智能AI Stable Diffusion 的论文介绍了从文字创建图像的技术。 Lay out phrasal verb. If you lay out a group of things, you spread them out and arrange them neatly, for example so that they can all be seen clearly. Grace laid out the knives and forks at the lunch-table. 人工智能AI does not process images through combining existing pictures人工智能AI 不通过组合现有图片来处理图像It is often misunderstood, but image generative 人工智能AI does not generate images through techniques like combining existing pictures. It learns how to draw pictures based on a vast amount of images and generates new images using words and image processing techniques.它经常被误解，但图像生成 人工智能AI 并不是通过结合现有图片等技术来生成图像的。它学习如何根据大量图像绘制图片，并使用文字和图像处理技术生成新图像。 Three steps of image generation生成图像的三步 There are three main steps in the 人工智能AI image-generation process. The first is "text conversion." This turns text inputs into representations that are easy for the 人工智能AI to interpret. The second is "image generation." This gradually makes the image closer to the input text-related image by removing the noise from an otherwise featureless image of pure noise. The third is "decoding." The image data, which had been compressed to enable the computer to perform quick calculations, is converted back into a form that is easy for humans to see, and the image is generated.人工智能图像生成过程分为三个主要步骤。第一个是“文本转换”。将输入文本转换为易于 人工智能AI 解释的表示形式。二是“图像生成”。从原本没有特征的纯噪声图像中去除噪声，使图像逐渐更接近输入文本。三是“解码”。为使计算机能够进行快速计算（而生成的图像数据），（现在）将这些被压缩的图像数据转换回人类易于查看的形式，并生成图像。 1 ) Convert input text转换输入文本One of the most attractive features of image generative 人工智能AI is the ability to create images by entering any text inputs. The text inputs are called the "prompt." In the first step of image generation, these text inputs by humans are converted into "vectors," quantified representations that are easy for the 人工智能AI to interpret. For example, the "cat" vector enables an 人工智能AI to quickly find features in images of cats.图像生成 人工智能AI 最吸引人的功能之一是能够通过输入任何文本输入来创建图像。文本输入称为“提示”。在图像生成的第一步中，人类输入的这些文本被转换为“向量”，即易于 人工智能AI 解释的量化表示。例如，“猫”向量使 人工智能AI 能够快速找到猫图像中的特征。 2 ) Denoising去噪 The second step is to generate images. Whereas humans paint on a blank canvas, 人工智能AI paints from images of pure noise filled with random data. The process of gradually removing noise from the image is repeated many times to generate the image based on the text inputs converted in the first step.第二步是生成图像。人类在空白画布上作画，而 人工智能AI 从充满随机数据的纯噪声图像中作画。重复多次从图像中逐渐去除噪声的过程，根据第一步转换的文本输入生成图像。 The process of removing noise brings the image closer to the text input. On the vast map of texts inside the 人工智能AI, the text representation created in the first step is set as the destination. As the 人工智能AI estimates how to denoise to generate an image that more closely resembles the input text representation, the image is gradually brought closer to the destination.去除噪声的过程使图像更接近文本输入。在 人工智能AI 内部庞大的文本地图上，第一步创建的文本表示被设置为目的地。随着 人工智能AI 估计如何去噪以生成更接近输入文本表示的图像，图像逐渐接近目的地。 resemble /rɪˈzɛmbl/ verb 3rd person present: resembles have a similar appearance to or qualities in common with (someone or something); look or seem like. "some people resemble their dogs" Similar:look like 、be similar to 3 ) Decoding for the human eye人眼解码In the third step, the image generated in the first and second steps are decoded and output as an image that the human eye can parse. In this visual explanation, we used images that are easy for the human eye to see, but 人工智能AI actually processes images that are smaller and more compressed in a space that machines can recognize, called "latent space". The use of small images makes it possible to generate images by a "shortcut", so to speak, and a high-performance computer can generate an image in just a few seconds after the text input is entered. Finally, the compressed data is converted back into an image consisting of color, width and height, resulting in a finished product.第三步，对第一步和第二步生成的图像进行解码，输出为人眼可以解析的图像。在这个视觉解释中，我们使用了人眼容易看到的图像，但 人工智能AI 实际上在机器可以识别的空间中处理更小、更压缩的图像，称为“潜在空间”。小图像的使用使得通过“快捷方式”生成图像成为可能，可以说，高性能计算机可以在输入文本后的短短几秒钟内生成图像。最后，将压缩数据转换回由颜色、宽度和高度组成的图像，从而产生成品。 人工智能AI learns in "reverse"人工智能AI“逆向”学习人工智能AI can determine the noise that should be eliminated for image generation because it has trained the difference between a image with noise and a image without noise. In the training process, various noises are artificially added to a clean image in the "reverse" direction of the generation process.人工智能可以确定生成图像时应该消除的噪声，因为它训练了有噪声图像和无噪声图像之间的差异。在训练过程中，在生成过程的“反向”方向上，人为地将各种噪声添加到干净的图像中。 Next, the 人工智能AI is trained to be able to generate an image before noise is added from an image with noise added. Through repeated training based on a vast amount of images, it is able to generate photo-realistic images even from images of pure noise.接下来，对 人工智能AI 进行训练，使其能够在添加噪声的图像中生成添加噪声之前的图像。通过基于大量图像的反复训练，即使是纯噪声图像也能生成逼真的图像。 This 人工智能AI mechanism is known as the "diffusion model." Stable Diffusion has been released as a pre-trained model, but derivative models that have received additional training by others are also available.这种人工智能机制被称为“扩散模型”。Stable Diffusion 已作为预训练模型发布，但也可以使用经过其他人额外训练的衍生模型。 The keys to training are "quantity" and "quality"训练的关键是“量”和“质”Both quality and quantity are essential in training data if an 人工智能AI is to generate elaborate images. For example, Stable Diffusion has trained approximately 2.3 billion images from a huge dataset that houses a large number of images and descriptive text collected from the Internet. The collection of the Metropolitan Museum of Art in the U.S., one of the largest museums in the world, contains more than 2 million items. 人工智能AI learns image features from more than 1,000 times that amount of data.如果人工智能要生成精美的图像，那么质量和数量对于训练数据都是必不可少的。例如，Stable Diffusion 已经从一个庞大的数据集中训练了大约 23 亿张图像，该数据集包含从 Internet 收集的大量图像和描述性文本。美国大都会艺术博物馆是世界上最大的博物馆之一，其藏品超过 200 万件。人工智能从 1000 多倍的数据中学习图像特征。 Just as humans develop an aesthetic sense through exposure to high-quality paintings, 人工智能AI needs to be trained with high-quality data. Accurate predictions are impossible if the images used for training are inconsistent with the explanatory text, if the image quality is poor or if the content is biased.正如人类通过接触高质量的绘画来培养审美感一样，人工智能也需要用高质量的数据进行训练。如果用于训练的图像与解释性文本不一致、图像质量差或内容有偏见，则不可能进行准确的预测。 With the August 2022 public release of the image generative 人工智能AI "Stable Diffusion," innovation in image generation 人工智能AI is advancing at an astounding rate. Image-generation 人工智能AIs have their own quirks based on their training data and mechanisms. While some images seem to captivate by capturing the entire history of human painting, other somewhat unusual images unintentionally leave people viewers laughing.随着 2022 年 8 月公开发布图像生成 人工智能AI“Stable Diffusion”，图像生成 人工智能AI 的创新正在以惊人的速度推进。基于其训练数据和机制，图像生成 人工智能AI 有自己的怪癖。虽然有些图像似乎通过捕捉人类绘画的整个历史而着迷，但其他有些不寻常的图像无意间让人们发笑。 Some may feel unease that 人工智能AI has stepped into the fundamentally human creative activity of "painting." However, there are many examples where technological innovation has expanded the horizons of human creativity. The technology of photography that emerged in the 19th century was a major catalyst for the creation of Impressionist paintings. In the world of chess and shogi, hybrid players like shogi champion Sota Fujii are employing new styles of play that were developed using 人工智能AI research.有些人可能会对人工智能涉足“绘画”这一人类基本的创造性活动感到不安。然而，有许多例子表明技术创新扩大了人类创造力的视野。19 世纪出现的摄影技术是印象派绘画创作的主要催化剂。在国际象棋和将棋的世界中，像将棋冠军藤井创太这样的混合棋手正在利用 人工智能AI 研究开发的新玩法。 What is now required of us is a proper understanding of how 人工智能AI works and a blueprint for coexistence between 人工智能AI and humans.现在需要我们正确理解人工智能的工作原理，以及人工智能与人类共存的蓝图。 如果您喜欢本文，点个“在看”再走吧。 来源:“高等数学上册” 声明:转载此文章仅供网友阅读了解该主题,出于传递更多信息之目的,并不意味着赞同其观点或证实其描述! |
