動漫真人化! 葬送的芙莉蓮 Midjourney教學 Prompt影片化 AI sora

蘋果妹
16 Mar 202414:17

TLDRThis tutorial showcases the process of using Midjourney to bring anime characters to life through AI drawing. The video focuses on three characters, emphasizing the importance of character selection, background analysis, and prompt refinement. It details the steps of using the /describe tool for inspiration, adjusting settings for anime-style images, and fine-tuning the prompts for a more realistic portrayal. The presenter shares tips on avoiding common AI pitfalls, such as generating too thin characters, and highlights the use of --s threshold for stylization. The video concludes with the presenter's satisfaction with the final images, encouraging viewers to experiment with Midjourney and share their creations.

Takeaways

  • 🎨 Use Midjourney to bring anime characters to life with a step-by-step process.
  • 🌟 Choose a character with simple features like Faerun for beginners to start with.
  • 📸 Utilize the /describe tool to reverse-engineer prompts from reference images.
  • 🎭 Switch to Niji mode for generating anime-style images and adjust settings as needed.
  • 🖌️ Refine prompts by removing irrelevant words and focusing on desired character traits.
  • 🌈 Test different prompt combinations to find the closest effect to the desired outcome.
  • 👗 Describe clothing and accessories in detail to achieve a realistic portrayal.
  • 📝 Use specific adjectives to convey the character's emotions and actions.
  • 🔄 Employ --s (Stylize threshold) to adjust creativity levels in the generated images.
  • 🚫 Use the -No command to avoid unwanted elements in the final image.
  • 🔍 Seek inspiration and descriptions from external sources like ChatGPT for complex costumes.
  • ✂️ Use image editing software like Photoshop for final touches and detail adjustments.

Q & A

  • What is the main topic of the video?

    -The main topic of the video is teaching viewers how to use Midjourney to bring their favorite anime characters to life in a realistic manner.

  • Which anime character is used as a demonstration in the video?

    -The anime character used as a demonstration in the video is Fulian.

  • What are the three steps involved in using Midjourney to create a realistic anime character?

    -The three steps are: 1) using cartoon pictures to reverse the Prompt, 2) refining the real-life Prompt, and 3) refining details locally after obtaining a finished picture.

  • Why is it important to avoid making the character look thin or yandere in the Prompt?

    -It is important because the character Faerun is described as not being thin and not fitting the stereotype of a yandere character. The goal is to create a realistic representation that stays true to the original character design.

  • What does Niji mode do in Midjourney?

    -Niji mode is a model specially used to generate anime pictures, which is essential when working with anime characters.

  • How does the --s parameter affect the generated images?

    -The --s parameter adjusts the Stylize threshold, with higher values leading to more creative and potentially confusing results, while lower values make the images closer to the given Prompt and instructions.

  • What is the significance of the 'White Conservative Dress, High Necked' in the Prompt?

    -This phrase is used to guide Midjourney in creating a more conservative and less revealing outfit for the character, avoiding unnecessary emphasis on physical attributes.

  • Why is the 'No anime' command used at the end of the Prompt?

    -The 'No anime' command is used to avoid generating images in an anime style, ensuring a more realistic and life-like portrayal of the character.

  • What is the purpose of the --Repeat tool in Midjourney?

    -The --Repeat tool allows users to send the same Prompt multiple times, which can be useful for refining the results and finding the most satisfactory image.

  • How does the Very Region tool help in refining the generated images?

    -The Very Region tool enables users to modify specific parts of the generated image, such as adjusting the hair or changing the color of clothing details, to better match the desired outcome.

  • What alternative method is suggested for handling high detail characters?

    -For characters with high detail, the video suggests using Stable Diffusion as it may be more suitable for capturing intricate costume and accessory details.

Outlines

00:00

🎨 Bringing Anime Characters to Life with Midjourney

The video begins with an introduction to using Midjourney, an AI drawing tool, to create real-life versions of anime characters. The presenter chooses Faerun from the anime Frillian as the subject for demonstration. The process is outlined in three steps: selecting a character, refining the prompt, and adjusting details. The first step involves using the /describe tool for inspiration, followed by switching to Niji mode for anime-specific generation. The second step requires returning to the original V6 setting and using a combination of image and text prompts to generate a real-life version of the character. The presenter emphasizes the importance of avoiding stereotypes and capturing the character's unique traits.

05:02

🌟 Refining the Real-Life Version of Faerun

This paragraph delves into the specifics of refining Faerun's real-life depiction. The presenter discusses the challenges of Midjourney's default settings, such as generating thin characters, and shares the solution of using descriptive words like 'chubby' and 'slightly round face' to achieve a balanced portrayal. The importance of clothing description is highlighted, with the use of 'white conservative dress' and 'high necked' to guide the AI's output. The paragraph also covers the addition of action, emotion, and environment to the character's depiction, aiming for a more dynamic and contextual portrayal.

10:03

🌈 Crafting Fulian's Image with Detailed Prompts

The final paragraph focuses on the creation of Fulian, another character from Frillian, with a detailed discussion on refining prompts for complex costumes and poses. The presenter shares the strategy of using the Describe tool for initial inspiration and the importance of specifying poses and clothing details, such as 'lying in the water' and 'high double ponytail.' The paragraph also introduces advanced tools like --Repeat for sending the same prompt multiple times and Very Region for局部调整. The presenter's experience with Photoshop for final touches is mentioned, along with a call to action for viewers to share their own Midjourney creations in the comments.

Mindmap

Keywords

💡Midjourney

Midjourney is an AI-based tool used for creating images, often utilized in the context of generating art or transforming existing images. In the video, it is the primary tool used to bring anime characters to life, with the presenter guiding viewers through the process of using Midjourney to create realistic versions of their favorite characters.

💡Anime Character

An anime character refers to a character from Japanese animated media, often stylized with distinctive features such as large eyes, exaggerated expressions, and unique hairstyles. In the context of the video, the presenter is teaching viewers how to use Midjourney to create lifelike versions of these characters, focusing on specific traits and characteristics to ensure accuracy.

💡Reverse Prompt

A reverse prompt is a technique used in AI image generation where an existing image is uploaded to the AI to analyze and generate a prompt based on that image. This is used to get inspiration for creating new images and to ensure that the generated content is similar to the original image in certain aspects.

💡Niji Mode

Niji mode is a specific setting within the Midjourney tool that is designed for generating anime-style images. It is a specialized model that tailors the output of the AI to create images that are stylistically similar to anime, which is particularly useful when the goal is to create realistic versions of anime characters.

💡Real-life Version

A real-life version refers to the process of transforming a fictional character or image into one that appears realistic and lifelike. This often involves adjusting the character's features, clothing, and overall aesthetic to resemble a real person or situation, which is a key objective in the video when using Midjourney to create anime character representations.

💡Stylize Threshold

The stylize threshold, denoted by the --s parameter in Midjourney, is a setting that adjusts the level of stylization in the generated images. A higher value results in more creative and stylized images, while a lower value yields images that are closer to the original prompt and more conservative. It is a crucial aspect of fine-tuning the output to match the desired aesthetic.

💡Prompt

In the context of AI image generation, a prompt is a set of descriptive words or phrases that guide the AI in creating an image. It includes both text and image inputs that provide the AI with the necessary information to generate the desired output. Prompts are essential in communicating the creator's vision to the AI, and they are meticulously crafted in the video to achieve the most accurate representation of the anime characters.

💡Fulian

Fulian is a character from an anime mentioned in the video. The presenter uses this character as an example to demonstrate the process of bringing anime characters to life using Midjourney. The character's specific traits, background, and costume details are discussed to illustrate how to create a real-life version that captures the essence of the original character.

💡Image Refinement

Image refinement is the process of adjusting and improving the generated image to better match the desired outcome. This can involve altering specific elements, such as hair, clothing, or facial features, and may require the use of additional tools or software, like Photoshop, to achieve a more realistic or accurate representation of the character.

💡Very Region

Very Region is a tool within Midjourney that allows users to modify a specific part of an image. This feature provides the ability to make localized adjustments to the generated content, such as changing the color of clothing or adjusting the style of hair, to better align with the creator's vision.

Highlights

The tutorial focuses on using Midjourney to bring anime characters to life.

The anime 'Fulian' is used as a case study for the demonstration.

Three characters from 'Fulian' are selected for the demonstration due to their higher probability of success.

The process begins with choosing a character, starting with Faerun due to his simple design.

Phelan's character analysis includes her age, appearance, and personality traits.

Faerun is described as a rare magic wizard and a super genius.

The tutorial emphasizes avoiding the portrayal of Faerun as thin or yandere in the prompts.

The process is divided into three steps: reverse prompting, refining prompts, and local detail refining.

The /describe tool is used to find inspiration for prompts by analyzing uploaded images.

Niji mode is a specialized model for generating anime pictures.

The importance of switching back to the original V6 setting for generating real-life versions.

The use of picture prompts combined with text prompts to refine the results.

The challenge of avoiding Midjourney's default to generate thin girls and the strategy to counter it.

The addition of prompts for actions, emotions, and environment to enhance the character depiction.

The use of --s to adjust the Stylize threshold for more creative or conservative outputs.

The demonstration of creating a real-life version of Fulilian with her simple pajamas and river scene.

The strategy for dealing with complex costumes and the use of ChatGPT for descriptions.

The final result showcases the transformation of anime characters into realistic depictions.