Stable Diffusion

In a blog post yesterday (October 22), Stability AI announced the release of Stable Diffusion 3.5, which marks a significant advancement in open source AI graphical models. Stable Diffusion 3.5 is available in Medium (released on October 29), Large and Large Turbo sizes, designed to meet the different needs of scientific researchers, enthusiasts, startups and enterprises, with the following introduction: Stable Dif...

AI painting, can be said to have been quite mature Closed source model Midjourney, easy to use, can produce photographic works, the effect is awesome. It is a few dozen knives per month, a little bit of money, and there is poor scalability. Character scene consistency requirements, or want to use the workflow to do exclusive painting tool applications, open source model is YYDS, such as stable diffusion series, and the recent fire Flux, media platforms have been slaughtered by the two lists. The charm of open source, lies in the scalability, controllability, but also with workflow packaged products, and the current effect is constantly approaching Mid...

Reddit users have reportedly been paying for nude photos of a fake, AI-generated woman. The photos, created by the AI tool Stable Diffusion, are extremely realistic looking and indistinguishable. The images first appeared on several Reddit forums a few months ago, garnering a lot of attention and comments from users. A photo of a beautiful woman generated by Stable Diffusion One user even commented after seeing Claudia's photo: "You're beautiful!" Others said, "Wow, that's really beautiful!" ...

I found that there are still many of my friends who have not been exposed to AI painting before. In order to help you quickly understand AI painting, as soon as possible to cross the door into the AI painting, start the AI painting journey, I personally feel it is very necessary to write an article for beginners to start. Of course, not only for beginners, there are some partners belong to the computer configuration is not enough to build the local SD environment Do content partners, not too concerned about the tool itself and technical capabilities, focus on creativity and content expression, can generate the effect I want to picture. For these types of partners, I will generally recommend a suitable for novice AI...

Today I would like to share with you a SDXL-based Lora model for clothing accessories classification background: classification background XUER. This Lora model was launched by the model training master Xuer in August. Currently, the V1 version is launched. The author claims that the image accuracy has not reached expectations, and the splitting is still a little different. It is being improved and optimized. I believe that a better version will be launched soon. However, even if the current version is not perfect, the effect is still very amazing, except for the inaccurate processing of clothing combination elements. Model download address (the network disk address at the end of the article can also be obtained) LiblibAI: https://ww…

Hands have always been a problem when using the AI painting tool Stable Diffusion. Unnatural or incorrect situations often occur when AI draws hands. Specifically, they are manifested in the following aspects. Missing or extra fingers: Sometimes, the hands of the characters in the AI-generated images may be missing some fingers, or there may be extra fingers. Distorted hands: The shape of the hand may look strange, such as the fingers are bent unnaturally, or the whole hand is twisted into an impossible shape. Wrong hand position: The hand may be placed in an inappropriate position on the character's body, such as the arm is crossed through the body...

How to use the Stable Diffusion tool has always been one of the questions that many friends often ask. Today, let's talk about how to experience the AI painting tool Stable Diffusion. In general, there are three main ways to experience the Stable Diffusion tool. Method 1: Local computer installation and deployment (computer configuration requirements) Method 2: Paid cloud platform (charged by traffic or time) Method 3: Free cloud platform (free, but limited functions) Related instructions, this article is mainly for SD...

Previously, I shared with you the relevant production methods of line drawing transfer. I will continue to share new practices about line drawing transfer in the future. Today, we will start to share another kind of picture transfer: illustration transfer. 1. Production method of picture to illustration [Step 1]: Selection of large model Here we recommend: Vientiane Melting Furnace | Anything V5/V3. Model download address (you can also get the network disk address at the end of the article) LiblibAI: https://www.liblib.art/modelinfo/1f26c86ea6a8442c856…

Regarding line drawing transfer, I have shared with you two ways to achieve text-generated images and image-generated images. Stable Diffusion [Application] [Line Drawing Transfer]: One minute to achieve image transfer and line drawing Stable Diffusion [Application] [Line Drawing Transfer]: One minute to achieve image transfer and line drawing without loar The previous two methods did not perform well in achieving line drawing transfer of architectural images. Today, I will share my personal practice specifically for line drawing transfer of architectural images. If you have good ideas for implementation, you can leave a message or add WeChat to communicate privately…

In the previous article, I shared with you how to achieve the Wenshengtu conversion of image line drawings. Stable Diffusion [Application] [Line Drawing Conversion]: One minute to achieve image conversion to line drawings Image line drawing conversion can also be achieved using the method of picture generation. Let's first look at the effect of using the picture generation method to achieve the image conversion to line drawings. 1. How to make a picture to line drawing [Step 1] Upload a white background image of the picture generation image Here we use Meitu Xiuxiu to make a pure white background background image. Upload this white background image in the picture generation function menu. [Step 2]: Selection of large models Here is recommended...

Today we start sharing the series of image-to-drawing in the application chapter. Image-to-drawing has many different application scenarios, such as image-to-line drawing, image-to-illustration, image-to-comic, etc. This is also one of the more common application directions of AI painting in actual scenes. Here we start with the production method of image-to-line drawing. 1. Production method of image-to-line drawing [Step 1]: Selection of large model Here we recommend: ReVAnimated, version v122. Model download address (you can also get the network disk address at the end of the article) LiblibAI: https://www.libl…

When using the Stable Diffusion tool for drawing, we often encounter this kind of problems, mainly manifested in (1) after a lot of drawing, we finally draw a picture that we are satisfied with, but there are some flaws in some parts of the picture, and the overall satisfaction of the picture can only reach 95%. (2) after a lot of drawing, we finally draw a picture that we are very satisfied with, and we want to get multiple pictures with similar effects. If we regenerate them, since SD generates pictures randomly by default each time, the elements of the newly generated pictures may not be what we want. For the above two scenarios,…

Today I will share with you a 2.5D model trained based on the Pony model: Speciosa 2.5D. Speciosa is a Latin word meaning "beautiful", so this model is good at generating semi-realistic cute Korean girls. The overall image quality of the pictures generated by this model is excellent, and the skin of the beauties is delicate. It can produce very high-quality pictures without the need for an amplification algorithm. On the official website, the author provides the training method of the model Steps #1: Js2Prony (v1.0) * 0.35 + DucHaiten-Pony-XL (pony-no-…

Recently I saw someone converting photos into line drawings in a book and the results were very good. I also referred to other people's methods and shared this technique with everyone. This method has two types: picture-based drawing and text-based drawing. Let me first explain the use of text-based drawing. Text-based drawing model: architecturerealmix This model can be downloaded on C station or Libu. Tips: BondyStudio, monochrome, greyscale, Dotted line, line shadow, exquisite, Lo…

Preface Students often ask which one is better, Midjourney or Stable Diffusion. In fact, it doesn't matter which tool is better. They each have their own strengths. In summary, SD is more suitable for fine control. If you have a sketch and need further processing, it is definitely better to use SD. But if you need to get creative inspiration quickly and efficiently, then definitely choose MJ. You have to know what your specific needs are, and then choose the right tool. Of course, students who have the conditions can learn both, which is obviously the best choice. This article provides a Midjourney...

Today I would like to recommend a very good general-purpose large model based on SDXL: SDVN7-NijiStyleXL. The drawing style of this model is very similar to the Midjourney style. The author combines animation, sketching, realism and other styles to create many great styles. We can use various artist style combinations supported by SDXL. The artist styles supported by SDXL can be found on the website: https://rikkar69.github.io/SDXL-artist-study/. This large model is similar to the realistic style…

In the hot summer, cool fruits are the most comfortable and cozy choice. Today I recommend a LORA model based on SD1.5: KK-Watery Fruit Photography. This model is a fruit photography model, which brings a watery special effect when the fruit falls, bringing a cool visual experience to the hot summer. The latest version of the model is V2.0. Some training materials are added to the V1.0 version, and the number of learning times and optimizer are adjusted. However, after actual testing, I personally feel that the effect of V1.0 is better. The author gives the recommended parameter settings on the official website: • …

Today I will share with you a two-dimensional model trained based on the Pony model: two-dimensional illustration. There are 4 different branch versions of this model. 1.5 version: loar model, recommended base model niji-anime two-dimensional 4.5. xl version: SDXL model version mix version: light and shadow weakened, reduce the SDXL version of the warm tone and a sense of overexposure. Pony version: based on the version trained by Pony Diffusion v6. The author's introduction to this model on the Liblib official website is relatively short, mainly related parameter recommended settings. Personal experience...

This tutorial mainly uses the tool Stable Diffusion to operate. How can we use Stable Diffusion to assist us in completing the model effect display? This tutorial is suitable for e-commerce design scenes, photography scenes and other practical uses of character design. The whole process is practical, and you need to absorb it slowly. After learning it, you can easily control the model's dress change. Let's go! Vincent map Libulibu generates it online. First use Vincent map, open the website, keep up with the rhythm, and start making https://www.liblib.art/s…

Today I would like to share with you a flat-style 2D model based on SD1.5: flat simple color flatanime-mix. For those who like flat style, this model is still very worth recommending. The latest version is V1.0, which is very friendly to the recognition of prompt words, and can express characters/scenes/objects. More often, this model can be used as a 2D base film in a flat illustration style. Model download address (you can also get the network disk address at the end of the article) LiblibAI: https://www.liblib.ar…

Stable Diffusion Quick Start: Image Generator This article provides a detailed image generator tutorial for Stable Diffusion beginners, guiding you on how to use the text-to-image function to start your AI art creation journey. Compared with Midjourney's image generator, the image generator in Stable Diffusion is much more powerful. In addition to image control, it also includes graffiti, partial redrawing, graffiti redrawing, uploading redraw masks, and batch processing. 1. Image generator interface introduction 2. Image generator first experience We can use the image material...

1. Introduction to the Wenshengtu function The Wenshengtu function refers to the process of converting the input text description into an image. Through the Wenshengtu function of Stable Diffusion, we can convert the creativity and imagination in our minds into specific images, providing more sources of inspiration for designers and artists. 2. Wenshengtu operation steps 1. Prepare text description First, we need to prepare the text description to be converted into an image. The more detailed and specific the description is, the more the generated image will meet our expectations. We prepare a prompt word here: Portrait, a young…

Today I would like to share with you a large model based on SD1.5 with a camera theme: Dreams and bubbles. This version mainly includes the following improvements: (1) The character poses and picture composition are richer, more dynamic, and more visually attractive; (2) The character composition and natural language understanding are both ideal. Adding (8k, RAW photo, best quality, masterpiece:1.2,) to the prompt words will greatly improve the output effect (3) Recommended parameter settings sampler...

[Realistic Photography] Dreams and bubbles L Introduction Today we introduce a LoRA model from the Dream Building Industry series: [Realistic Photography] Dreams and bubbles, which is a realistic photography model with Asian faces. The model has richer, more dynamic and visually attractive character postures and picture composition, and has good drawing effects in character composition and natural language understanding. It also includes excellent viewing effects, such as night and low-light photography, portrait photography, background blur effects, etc. It can directly produce the photography effects of young and beautiful Asian girls, such as students, car models, etc. For…