Illustrating Children's Books with Midjourney AI

Table of Contents

1. Introduction
2. Using Chat GPT to Create a Story
3. Using Mid-Journey AI to Generate Images
4. Getting Started with Mid-Journey AI
5. Generating Images with Mid-Journey AI
6. Advanced Image Prompting Techniques
7. Challenges with Artificial Intelligence
8. Compiling Images into a Storybook with Canva
9. Conclusion

Using Mid-Journey AI to Generate Images

In the previous video, we covered how to use Chat GPT to assist in creating a story. In this video, we will be discussing how to use Mid-Journey AI to support You in creating graphics that will bring your story to life. Mid-Journey AI is one of the oldest image-generating AI out there and has a very strong reputation. As of recording this, Mid-Journey AI has over 1 million active users running on Discord, making it the largest Discord server in use currently.

Getting Started with Mid-Journey AI

To get started using Mid-Journey AI, you will need both a Discord account and a Mid-Journey account. Start by going to discord.com. From there, you can sign in, create an account, and download Discord for your computer if you would like. Discord.com has a nice beginner's guide to Discord where it will walk you through what is a server and how to do direct messaging via Discord.

Next, you will need to create a Mid-Journey account. Go to Midjourney.com and then create your account. Once you have your Discord and Mid-Journey account set up, go to Discord and search for the Mid-Journey server that will pop up on your left, and it will look something like this:

[Insert Image Here]

Generating Images with Mid-Journey AI

When generating images in Mid-Journey, there are a few ways to do this. If you want to do this for free, you can open up one of these newbie rooms. From here, you can put in your Prompts to generate images. The downside of doing this is that your images will be filtered in between other images that are on the platform. The upside of doing this is that this is a free way to generate images.

I chose to pay thirty dollars to upgrade my Mid-Journey account, and that allows me to set up a direct message with the Mid-Journey bot where all of my images that I'm generating are in one constant stream. I'm not having to look through the public stream of images when trying to generate image images.

You start by typing forward slash and then imagine. Hit the tab button, and then you start with your prompt. Generating prompts is a skill that takes some practice. Keeping in mind that you are telling a computer what Type of images you want to create, the computer will do its best to interpret your intentions to create an image that fits your needs.

I am going to start by looking back at my chapter story and look at my chapter outlines and try to generate images Based on that. In my first image, my character Jace is going to find an old compass in his grandfather's attic. I'm going to try to describe my character as best as possible to the Mid-Journey bot.

When you are done with your prompt, hit the enter button. We Are going to use the consistent prompt of a six-year-old light-skinned black boy with short curly hair, Jace Thompson, smile on his face, comic style. That will be our base prompt for just about every image we try to generate that has our character in it.

A tip for generating images in Mid-Journey is to always use a consistent style. I give my character a name because that is one way to train the AI to associate different characteristics with a specific prompt.

You can see the artificial intelligence is slowly rendering four different images. The percentage here is how much of the image is complete. Once the Mid-Journey bot is done rendering your images, you can click on it to see a larger Scale version.

At this point, I want to point out the different buttons that Show up underneath the images that were generated. You see an entire row of views and a row of V's. The images are numbered one, two, three, four in that order. If I press the U button, what that will do is it will upscale an image.

I kind of like this image in the bottom right, which is the fourth image. So, I'm going to say I want to upscale that fourth image. I also would like to regenerate all these images and see if I get something else that I like. So, I'm going to hit that regenerate button as well.

V1 through V4 stands for variation. What variation will do is it will take your base image. If I click to V4, it will take this image and create four new variations of that image that look pretty similar. You can see here that I have four different images that were regenerated from my original prompt. I do like image 2 and image 4 as well, so I'm going to upscale them, and I'm also going to regenerate my original prompt again.

There is no harm in generating images. You can generate as many as you would like. I would encourage doing this to find images that you think best suit the need for your book.

Advanced Image Prompting Techniques

A couple of advanced image prompting techniques are to use a prior image to base your prompt off of. The AI will try to use features from that first image in your Second image and create some consistencies there. Another advanced tip with prompting is using seeding. When trying to have images that look largely the same, you can use an upscaled version of an image with the little envelope button in the top right as a reaction, and then you can grab its seed number and enter that at the end of your image prompt.

Challenges with Artificial Intelligence

As with any artificial intelligence, using Mid-Journey to create images is not perfect. A couple of things that I have found that artificial intelligences really struggle at are creating faces that are similar from picture to picture to picture. This takes a lot of work to generate similar-looking characters from page to page. Another difficulty that I've found with artificial intelligence is creating hands. You will oftentimes find that a hand has more digits on it than it should or that the hands are distorted in some funny way. I would recommend practicing some basic photo editing skills so that you can touch up things like characters' hands.

Compiling Images into a Storybook with Canva

Now that you have all of your images created, the next question is how do you compile them into a book format? For that next step, click here on how to use Canva to create a beautifully constructed storybook. That is where you will really see your book take Shape.

Conclusion

In conclusion, Mid-Journey AI is a powerful tool that can help you generate images for your children's book. With a little bit of practice and patience, you can create images that bring your story to life. Remember to use a consistent style and to keep trying different prompts until you find the images that best suit your needs.