2024最新版midjourney以图生图进阶教程,mj生成相似图片,v6 圖生圖,以图生图的具体步骤,以图生图参数iw权重讲解,iw怎么用

小鱼儿AI学院
24 Mar 202414:22

TLDRIn this tutorial from Xiaoyuer's AI Academy, viewers are guided through the process of generating images from existing images using midjourney's V6 version. The video emphasizes the importance of consistency in character design for storytelling and offers a step-by-step guide on how to upload a reference image, craft a descriptor, and use parameters like -iw for image weighting to enhance similarity to the original. The host also discusses the cost-effectiveness of midjourney, suggesting shared rental options for beginners and showcasing the impact of different IW weights on the generated images. The summary concludes with a comparison of the effects produced with and without the IW parameter, highlighting the improved results when using an IW weight of 1.5 for greater similarity to the original image.

Takeaways

  • 🎨 **Using Images to Create Images**: Xiaoyuer teaches how to generate images from existing images, maintaining character consistency which is crucial for storytelling or animations.
  • 📚 **Basic Knowledge Required**: For beginners, it's recommended to start with Xiaoyuer's basic tutorials on midjourney to understand the foundation before proceeding.
  • 🌐 **Discord Registration**: Starting with midjourney involves registering on Discord and joining a server dedicated to midjourney to learn from scratch.
  • 🖼️ **Uploading a Reference Picture**: The process begins by uploading a reference picture, which can be a self-made image or an existing one.
  • 📝 **Descriptor Entry**: A descriptor, such as 'A super cute, super fat cat wearing a dress', is entered to guide the image generation process.
  • 💻 **Midjourney Interface**: Users interact with the midjourney interface to input commands and parameters for generating images.
  • 🔄 **Retrying Failed Deliveries**: If an error occurs, such as a delivery failure, users can retry the process by re-entering the parameters.
  • 💲 **Pricing Plans**: Midjourney offers different pricing plans, including a paid standard plan, which may not be cost-effective for novices.
  • 🏡 **Shared Rentals**: For cost-conscious users, midjourney provides shared rental options that are more affordable and suitable for those just starting.
  • 🔗 **Image Linking**: In the process, linking an image directly to the prompt can be done by copying and pasting the image link into the command.
  • ⚖️ **IW Parameter**: The Importance Weight (IW) parameter is used to maintain the consistency of generated images, with values ranging from 0.5 to 22.
  • 🔍 **Zooming In**: Xiaoyuer demonstrates the effectiveness of adding IW weight by zooming in on the generated images to show the improved similarity to the original.

Q & A

  • What is the main topic of Xiaoyuer's AI Academy course?

    -The main topic of the course is how to create images from existing images using midjourney, with a focus on maintaining character consistency in story creation.

  • What is the first step in the midjourney image creation process?

    -The first step is to upload a reference picture, which can be either an existing image or one created by the user.

  • How can a user generate a descriptor for their reference picture?

    -A user can use Google Translate to input the descriptor they want, or use GPT to generate a descriptor for them if they are creating a story.

  • What is the purpose of the IW parameter in the midjourney image creation process?

    -The IW parameter is used to maintain the consistency of a character when generating images, with values ranging from 0.5 to 22, where a higher value indicates greater similarity to the original image.

  • What does the V6 version in the script refer to?

    -The V6 version refers to the specific version of midjourney being used to generate images, which may have different features or capabilities compared to other versions.

  • How does the user know if they need to retry an image generation in midjourney?

    -If there is a delivery failure, midjourney will display an error message, and the user can click a retry option to attempt the image generation again.

  • What are the different pricing plans mentioned for midjourney?

    -The pricing plans mentioned include a paid plan and a standard plan, with monthly costs ranging from 30 'knives' (a term used in the script, possibly referring to a currency), which is more than 200 RMB.

  • How can a novice afford to use midjourney for image creation?

    -Novice users can opt for a shared rental plan, which is a more cost-effective way to access midjourney's services, with prices starting at a standard 6-person monthly payment of 59.

  • What is the significance of the AR parameter in the midjourney command?

    -The AR parameter is used to set the aspect ratio of the generated image, which can be important for ensuring the image fits within certain dimensions or looks as intended.

  • How does the user know which image they have generated is the one they want to use?

    -After generating images, the user can review them and select the one they are most satisfied with, for example, by clicking on an image they find excellent.

  • What is the advice given for users who are confused by the midjourney process?

    -Users who are confused are advised to go back to Xiaoyuer's channel and watch some basic tutorials that cover the use of basic commands, aspect ratios, and other foundational knowledge.

Outlines

00:00

🎨 Creating Pictures from Existing Images

Xiaoyuer's AI Academy introduces a method for creating pictures from existing images to maintain character consistency in drawings or story videos. Xiaoyu demonstrates the process using a set of pictures made from pictures, emphasizing the importance of using reference images and descriptors. The video guides viewers on how to upload a picture, use Google Translate for descriptors, and input parameters for the AI to generate images. It also discusses the cost of using midjourney tools and suggests affordable rental options for beginners.

05:02

🛒 Developing a Continuous Character with AI

The script explains how to develop a continuous character using AI by changing prompt words and using image links as references. It details the process of creating a shopping scene with a cat pushing a shopping cart in a supermarket. The importance of using the correct parameters and the IW (image weight) parameter for generating images with higher similarity to the original is highlighted. The video also provides tips on adjusting the IW weight for better results and suggests viewing basic tutorials for a better understanding of the process.

10:05

🔍 Enhancing Image Similarity with Weight Parameters

This paragraph focuses on enhancing the similarity of generated images to the original by adjusting the IW weight parameter. Xiaoyu shares personal opinions on the effectiveness of the blur-to-clear transition technique and the impressive results from Midjourney. The video shows comparisons between images generated with and without the IW parameter, demonstrating the impact of this adjustment on similarity. It concludes with a summary of the process for using pictures to generate similar pictures, emphasizing the use of image links and weight parameters to achieve the desired outcome.

Mindmap

Xiaoyuer's AI Academy
讲师介绍
教授如何使用图片生成图片
保持角色一致性
课程目的
课程介绍
Discord 注册
零绘画基础学习
midjourney 基础教程
基础知识
使用自己制作的图片
使用 Google Translate 输入描述
上传参考图片
示例描述:穿裙子、拿时尚包的超可爱胖猫
输入描述符
V6 版本参数
图像宽高比 (AR)
图像生成命令 (/slash image)
使用参数
IW 参数范围:0.5 到 22
IW 参数作用:保持图像一致性
IW 参数权重
操作步骤
标准计划与付费计划
月付价格
midjourney 付费计划
共享租赁选项
独立租赁选项
租赁方案
成本与付费
使用图片链接与添加 IW 参数的比较
IW 参数对相似度的影响
生成效果比较
使用相同的图片和不同的描述符
场景变化对生成效果的影响
连续角色生成
图像生成效果
使用图片链接生成图片
添加 IW 参数以提高相似度
关键点回顾
订阅频道
观看其他视频教程
观众互动
教程总结
Midjourney 图片生成教程
Alert

Keywords

💡midjourney

Midjourney refers to a specific tool or process used for generating images from existing images or descriptions. In the context of the video, it is a platform or software that allows users to create new images by providing a reference image and a descriptive prompt. It is central to the video's theme as it is the primary method demonstrated for image generation.

💡V6 version

The V6 version is a specific iteration or version of the midjourney tool being used. It implies that there have been previous versions and that this one comes with its own set of features and improvements. The script mentions using the V6 version to achieve certain effects, indicating its importance in the process.

💡descriptor

A descriptor in this context is a text input that describes the characteristics or appearance of the image the user wishes to generate. It is a key component in the image generation process as it guides the AI in creating the desired output. For example, the script mentions 'A super cute, super fat cat wearing a dress, and carrying a fashionable bag is coming to us' as a descriptor.

💡IW weight

IW weight stands for Image Weight and is a parameter used in the midjourney tool to control the influence of the reference image on the generated image. A higher IW weight means the generated image will be more similar to the reference image. It is used to maintain the consistency of a character across different images, as shown when the presenter adds an IW weight of 1.5 or 2 to achieve a higher similarity.

💡prompt

A prompt is a form of input, usually a phrase or sentence, that is used to instruct the AI on what kind of image to generate. It is closely related to descriptors and is often used interchangeably. In the video, prompts are used to guide the creation of various images, such as 'The cat is sitting on the ground crying' or 'The cat walked into the supermarket with a bag'.

💡character consistency

Character consistency refers to the maintenance of a character's visual appearance across multiple images or scenes. This is important in storytelling or when creating a series of images featuring the same character. The video discusses using midjourney and IW weight to ensure that the generated images of the cat maintain a consistent look.

💡Google Translate

Google Translate is mentioned in the script as a tool that can be used to translate descriptors into different languages. This can be helpful for users who want to generate images with descriptors in a language they are not familiar with, thus broadening the accessibility of the midjourney tool.

💡GPT

GPT, or Generative Pre-trained Transformer, is referenced as a tool that can assist users in creating descriptors for their images. It is an AI language model that can generate human-like text based on given prompts, which can be useful for users who need help formulating their image descriptors.

💡image link

An image link is a URL that points to the location of an image on the internet. In the context of the video, image links are used as references for the midjourney tool to generate new images. The presenter discusses copying and pasting image links into the tool to create images that are similar to the ones found at those links.

💡shared rental

A shared rental in the context of the video refers to a subscription model where multiple users share access to the midjourney tool, splitting the cost. This is presented as a more cost-effective option for users who are new to the tool or do not require exclusive access.

💡AI services

AI services are artificial intelligence tools or platforms that provide various functionalities, such as image generation, language translation, or text generation. In the video, AI services like midjourney and GPT are discussed as tools that facilitate the creation of images based on textual descriptions and reference images.

Highlights

Xiaoyuer shares a tutorial on creating images from existing images for character consistency in story creation.

Introduction to midjourney, a tool for generating images from prompts and reference pictures.

Explanation of how to use a reference picture and create your own for generating images.

Demonstration of using Google Translate and GPT to generate descriptors for image creation.

Step-by-step guide on entering a descriptor and using parameters like -ar and --v for image generation.

Illustration of retrying a failed image generation attempt with midjourney.

Discussion on the cost-effectiveness of midjourney, including different pricing plans and rental options.

Introduction of AI services and tools commonly used with midjourney.

Explanation of how to choose and pay for midjourney services, including shared and individual plans.

Demonstration of selecting and creating a picture based on a generated image.

Tutorial on changing prompt words for different scenes and maintaining character consistency.

Importance of the IW parameter for generating images with higher consistency to the original picture.

Comparison of image generation effects with and without the IW parameter.

Adjusting the IW weight to achieve desired levels of similarity between generated and original images.

Use of blur-to-clear transition in image creation for higher similarity.

Recommendation to watch basic tutorials for beginners to understand the use of commands and aspect ratios.

Final summary and reminder to subscribe for new content and tutorials on midjourney image generation.