Forget Lensa AI! Make Avatars with Midjourney

Making AI Magic
21 Dec 202208:26

TLDRIn this video, Jen introduces viewers to Midjourney, an alternative to the popular Lensa AI app for creating personalized avatars. Unlike Lensa, which requires multiple photos to train its AI model, Midjourney can generate detailed avatars from just one selfie. Jen explains the process of using Midjourney, including selecting the right images, crafting a prompt, and utilizing the 'chaos' parameter for diverse outputs. She also highlights the benefits of Midjourney's higher quality images and its free trial, encouraging viewers to explore the platform's capabilities for creating a wide range of avatar styles.

Takeaways

  • 🎨 **Midjourney Creation**: Midjourney is a text-to-image generator that allows you to create personalized avatars in various styles based on your description.
  • 📱 **Simple Start**: Unlike Lensa AI, you only need one selfie to start creating avatars with Midjourney, though you can add more for better results.
  • 💡 **Realistic Output**: Midjourney's AI tends to produce more realistic avatars compared to the stylized output of Lensa AI's stable diffusion.
  • 🔍 **Control Over Content**: Midjourney offers better control over nudity and sexualized images, reducing the likelihood of inappropriate content.
  • 📊 **High-Quality Images**: Midjourney creates higher quality avatars that can be upscaled to 2048 pixels, compared to the 1024 pixel squares from Lensa AI.
  • 🛠️ **Customization**: With Midjourney, you're not limited to pre-selected styles; you can prompt for any character or style you desire.
  • 💰 **Cost-Effective**: Midjourney allows you to create 100 avatars for free, whereas Lensa AI requires a yearly membership fee after the initial avatars.
  • 🌐 **Versatility Beyond Avatars**: Midjourney is not limited to avatar creation; it can be used for a wide range of image generation purposes.
  • 📸 **Image Selection**: Choose clear, close-up selfies that represent you well, avoiding busy backgrounds or additional people that might confuse the AI.
  • 🔗 **URLs for Images**: Obtain URLs for your images by uploading them to Discord and copying the link, which Midjourney will then use in the image generation process.
  • ⚙️ **Prompt Structure**: Use a simple and descriptive prompt for Midjourney, including your gender, image links, and a brief character description for the best results.
  • 🔄 **Re-Roll for Variation**: If the initial results aren't satisfactory, use the re-roll function to generate more images until you find one you like.

Q & A

  • What is the main topic of the video?

    -The main topic of the video is about creating personalized avatars using Midjourney, an alternative to the Lensa AI app.

  • How does Lensa AI create avatars?

    -Lensa AI creates avatars by training a model with 10 to 20 photos of the user's face using stable diffusion AI.

  • What is the difference between Midjourney and Lensa AI in terms of avatar creation?

    -Midjourney is a text-to-image generator that creates avatars based on a description provided by the user, while Lensa AI requires multiple photos of the user's face to train a model for avatar creation.

  • What are the advantages of using Midjourney over Lensa AI?

    -Midjourney allows users to create avatars with more variety and control, does not require a subscription fee, and offers higher quality images that can be upscaled to 2048 pixels.

  • How does Midjourney handle nudity and sexualized images?

    -Midjourney has better controls than stable diffusion when it comes to nudity and sexualized images, making it less likely to produce inappropriate content.

  • What is the file size of images created by Midjourney?

    -Midjourney creates more detailed and higher quality avatars, with images that can be upscaled to 2048 pixel square.

  • How many avatars can you create for free with Midjourney?

    -With Midjourney, users can create 100 avatars absolutely free.

  • What is the process of creating an avatar in Midjourney?

    -The process involves providing a selfie, specifying gender, entering image links, describing the avatar, and using the 'slash imagine' command with version 4 of Midjourney to generate the avatar.

  • How can one upscale the generated avatar images in Midjourney?

    -Users can upscale the generated avatar images by selecting the U1, U2, U3, or U4 buttons in Midjourney.

  • What are some tips for choosing the right images for Midjourney avatar creation?

    -Choose close-up photos of your face without hands or other people, and remove busy backgrounds to allow the AI to clearly see and interpret your features.

  • How can one get more diverse outputs when creating avatars with Midjourney?

    -Adding 'dash dash chaos space 50' to the prompt will tell Midjourney to generate more diverse outputs.

  • What is the significance of using version 4 of Midjourney for avatar creation?

    -Version 4 of Midjourney is particularly good at combining images and prompts, making it the best choice for creating personalized avatars.

Outlines

00:00

🎨 Introduction to Lensa AI and Mid-Journey for Avatar Creation

This paragraph introduces the audience to Lensa AI, an app that generates avatars based on your selfies in various styles. Jen, the presenter, suggests that instead of just creating a drawing or painting of oneself, one can use these apps to become a different character entirely. She explains that Lensa AI uses stable diffusion to create personalized avatars from 10 to 20 photos of the user. Mid-Journey is introduced as an alternative text-to-image generator that can create avatars from a single selfie, with better control over potentially inappropriate content. The paragraph also highlights the differences between Lensa AI and Mid-Journey, such as the number of photos required, the style of output, and the file size quality. It concludes with a teaser about the process of creating avatars with Mid-Journey.

05:01

🚀 How to Use Mid-Journey for Creating Custom Avatars

The second paragraph delves into the specifics of using Mid-Journey to create custom avatars. It starts by discussing the process of selecting appropriate selfies for the AI to analyze, emphasizing the need for clear, close-up images without hands or busy backgrounds. The method for obtaining URLs for these images is explained, along with the process of entering a prompt in Discord to generate the avatar. The paragraph provides tips for creating a successful prompt, including using descriptive words and avoiding overly complex instructions. It also mentions the use of version 4 of Mid-Journey for better image and prompt combination. The audience is encouraged to experiment with different prompts and re-rolls to find the best avatar. The paragraph concludes with advice on how to refine the avatar creation process, including using additional details and the 'chaos' parameter for more diverse outputs, and how to upscale and download the final avatar for personal use.

Mindmap

Keywords

💡Lensa AI

Lensa AI is an application that has gained popularity for creating avatars based on the user's selfies. It uses a process involving stable diffusion to train a model that resembles the user's face, allowing it to generate various styles of avatars. In the video, it is compared with Midjourney, another tool for creating avatars, highlighting the differences in their approaches and offerings.

💡Midjourney

Midjourney is a text-to-image generator that can create personalized avatars based on a user's description. It is similar to stable diffusion but offers better control over the output, particularly regarding nudity and sexualized images. The video emphasizes that Midjourney allows for more creativity and customization, as users can specify various styles without being limited to a set number of styles provided by the app.

💡Stable Diffusion

Stable diffusion is an open-source AI technology used by Lensa AI to generate avatars. It trains a model based on multiple photos of the user to create avatars in different styles. The term is mentioned in the context of comparing the technical approach of Lensa AI with that of Midjourney, which also uses a similar but distinct technology.

💡Avatar

An avatar, in the context of the video, refers to a digital representation of a person, often used for online profiles or in virtual environments. The video discusses how both Lensa AI and Midjourney can create avatars, but with different methods and styles, allowing users to express their identity in various ways.

💡Profile Pics

Profile pictures, or 'profile pics,' are the images that users display on their social media or online accounts to represent themselves. The video script discusses how Midjourney can be used to create unique and personalized profile pictures in different styles, offering users more flexibility than traditional methods.

💡Text-to-Image Generator

A text-to-image generator is a type of software that creates images based on textual descriptions provided by the user. Midjourney is an example of such a generator, and the video explains how it can be used to generate avatars by interpreting the user's description and creating an image accordingly.

💡Dream Booths

Dream Booths is a concept mentioned in the video that refers to a method where a coder with the right computer setup can create personalized avatars using stable diffusion AI. It is presented as an alternative for those with technical skills looking to create avatars similar to what Lensa AI offers.

💡File Size

File size refers to the amount of digital space a file occupies. In the context of the video, it is mentioned in relation to the quality of avatars created by different apps. Midjourney is noted for creating more detailed and higher-quality avatars, which may result in larger file sizes compared to other apps.

💡Prompt

In the context of Midjourney, a prompt is a textual description given by the user to guide the AI in generating an image. The video provides a detailed explanation of how to structure prompts to create avatars in different styles, emphasizing the importance of simplicity and strong character descriptions for effective results.

💡Discord

Discord is a communication platform where Midjourney operates. Users can join the Midjourney beta through Discord and use it to create avatars. The video script provides instructions on how to use Discord for this purpose, including how to obtain URLs for images and how to structure prompts for avatar creation.

💡Chaos

In the context of Midjourney, 'chaos' refers to a parameter that controls the diversity of the output images. By adding '--chaos' followed by a number (e.g., 50) to the prompt, users can instruct Midjourney to generate more varied avatars. This feature is highlighted as a way to increase the range of styles and options available to the user.

Highlights

Lensa AI is a trending app that creates avatars in different styles based on your selfies.

Midjourney offers a more customizable avatar creation experience without the need for a yearly membership fee.

Midjourney is a text-to-image generator that can create personalized avatars based on a description.

Midjourney V4 excels at combining images and text, allowing for the creation of personalized avatars that resemble the user.

Unlike Lensa, Midjourney does not require multiple photos to train its AI model; one selfie is sufficient.

Midjourney's output tends to be more realistic and less stylized compared to Stable Diffusion.

Midjourney provides better controls over nudity and sexualized images than Stable Diffusion.

Midjourney avatars can be upscaled to 2048 pixels square, offering higher quality than many other apps.

With Midjourney, users are not limited by preselected styles and can choose from a wide range of styles, including punk rocker or Renaissance royalty.

Midjourney allows users to create 100 avatars for free, providing a cost-effective alternative to paid services.

Creating avatars with Midjourney involves providing a few selfies and using a simple prompt structure.

Photos used for avatar creation should be close-ups of the face without hands or busy backgrounds to avoid confusion for the AI.

Midjourney can generate avatars from just one image, but additional images can be added for more accurate results.

To create an avatar, users input a gender specification, selfie image links, and a descriptive prompt for the character they want.

Midjourney V4 should be specifically requested for the best results in combining images and prompts.

Users can upscale their preferred avatars using the U buttons and add additional detail to the prompt for more stylized or diverse outputs.

Midjourney is not only for avatar creation but is a versatile tool for a wide range of image generation applications.

The video provides tips and tricks for creating avatars in Midjourney and encourages viewers to share their discoveries.