[🔰For Beginners] AI Image Generation: Where to Start? How Starting with 'Character Creation' Worked for Me [Part 1] | Pollo.ai | AI Image Generation | Dark Fantasy

If we haven't met, nice to meet you!
To all my followers who always read my posts, hello, good afternoon, good evening!
Welcome, I'm Ryuto Natsume (*´ω｀)
I work as an AI image and video generation creator with a dark world view.

This article is a sponsored post for Pollo.ai (*´ω｀)

This article is an introductory guide [Part 1] to AI image generation using 'Pollo.ai'.

April, the season when you want to start something new.
"I'm curious about AI image generation, but it seems difficult..." I wrote this for those of you who feel that way.

Without getting into difficult topics, I will deliver my own super beginner-friendly guide.

In this first part, I plan to explain image generation, and in the second part, video generation using those images.
Why don't we experience the world of AI image and video generation together in this two-part series? (*´ω｀)

I am an eternal beginner, meaning I go about my daily activities with the mindset of "never forgetting one's original intentions."
This also serves as my own study, as it allows me to gauge whether I can explain things well enough to help readers take their first step into "AI image generation."

P-please, go easy on me (*´ω｀)✨

※This article is based on the creative experience I, "Ryuto Natsume," have had up to "now," and I hope you will accept it as one reference point, thinking, "there is this way of looking at things too."

Also, regarding how to create prompts, it varies from creator to creator.
It's perfectly fine if you just get a "feel" for it (*´ω｀)

📢 Table of Contents for this article [AI Image Generation]

① What is Pollo.ai

② What I value before creating prompts: "Concept"
→ Trying to create your own "character"

③ List of Pollo.ai's AI image generation features and introduction to the operation screen

④ Based on ② and ③, building a prompt and actually generating an image
(Creating a common prompt)
※Explaining each operation with screen screenshots

⑤ Conclusion

I intend to keep the content simple, so I would be grateful if you could just get a feel for what Pollo.ai is like (*´ω｀)

Well then, let's get started ╰(*°▽°*)╯

① What is Pollo.ai

Pollo.ai is an all-in-one AI image and video generation platform.

To put it in a bit more detail...

It is a versatile AI platform that combines various AI tools into one, which users can choose according to their needs (*´ω｀)

The HP looks like this.

Usually, AI image generation tools are separate services, requiring registration or fees every time you use them.

But with Pollo.ai,popular models like Midjourney, Stable Diffusion, FLUX, and Nano Banana are all gathered here!

And you canstart for free✨

Pollo.ai's 'generation' consumes what are called 'credits'.

Also, there are campaigns where those 'credits' are more than half off, so if you have an 𝕏 account, please be sure to follow and check it out (*´ω｀)

First, just registering for Pollo.ai is enough.
Let's dive into the world of AI image generation together from here ╰(*°▽°*)╯

② What I value before creating a prompt: 'Concept'

Well then, you might be full of enthusiasm, thinking, 'Alright! I'm going to generate images! ╰(*°▽°*)╯'
Yes, but everyone, when you think of AI image generation or AI video generation, don't you think of 'prompts'?

Prompts are certainly important.
But there is something you need before prompts.

That is the 'concept'.

…Well, this is just something I always keep in mind when using AI💦
I hope you take this as just a 'reference' (*´ω｀)

A prompt is a tool.
Before mastering the tool, do you have an image in your mind of 'what you want to create'?
(I think of it as a 'love letter' when I create it.)

AI gives you ideas.
But before that, you have to tell it what you want to draw.

AI accelerates your concept.
But the concept itself can only be born from you.

This is the first AI-generated image I created with Pollo.ai.

While I now theme my work around a 'dark worldview,' at the time, I couldn't step into such a niche world, and I was anxious about whether I could keep it up, so I just drew 'a woman and a cat' for the time being.

Yes, I set the 'theme' to 'a woman and a cat'.

'A woman and a cat'—everyone likes that, right?

Is it too cliché?

It doesn't matter what you start with.

😻 Let's try creating your own 'original character'

This is where you think a little further.

・'What kind of' woman? 'What kind of' cat? ← Their 'appearance'

・What are they doing? What happened? ← Laughing, hugging

At this point, you already have a 'concept'.
Once you've decided on a theme, the 'foundation' is already built, so it's like thinking about accessories—the 'personality'—to decide 'how to go about it' from there✨

When it comes to future 'image generation' or 'video generation', I think the easiest thing to do is to create your own 'character', that is.
Of course, it doesn't have to be a 'person' either (*´ω｀)✨

Here, the theme is 'dark'.
From there:
・A beautiful woman with silver-white hair (🤤 Oh ho, it's fine to exaggerate!!)
・Wearing glasses
・Purple eyes
・A cute black monster costume (depending on the AI, it might draw it as a hat and outfit separately like this)

This is how you devise a 'character'.
From there, AI image generation is already complete just by deciding 'what to do' with the character (*´ω｀)

Creativity requires 'personality' in any field.
Even just one icon can change the impression.

Especially on this Note platform, I've been seeing more 'AI image generation' for header images and article content—not just text, but 'image illustrations', 'diagrams', and 'eye-catchers'.

Your own character can become the 'face' of your Note articles and help liven them up.

That's right, the 'character' created with AI image generation becomes a 'topic' in itself✨
Plus, I think the visual effect is truly amazing for 'making an impression' that 'this is that person'.

Once you start generating images with AI, you'll begin to understand the characteristics of the AI, thinking, 'Ah, this person is using that AI'.
To gain that kind of 'knowledge', the best thing to do is to try it yourself first and get a feel for the AI's quirks (*´ω｀)

Now then, shall we try creating a 'character'—oh ho 🤤?

But before that, shall we take a look at what kind of image generation AI Pollo.ai has on board (*´ω｀)

➂ List of AI Image Generation on Pollo.ai

Pollo.ai AI Image Generation Model List

• Nano Banana (Google)

• Midjourney

• Recraft

• Ideogram

• Stable Diffusion

• FLUX

• Seedream

• Dall-E (Open AI)

• Imagen (Google)

• GPT-4o (Open AI)

• Qwen Image (Wan AI)

• Wan

• Grok Imagine (xAI)

• Kilig

• Pollo image

(　＾ω＾)...

Ah, I understand.
I completely understand how you feel.
I apologize for piling on, but now 'Video Generation AI' is being added to this as well.

...It's amazing, isn't it?

Here, I have picked out four AIs.
I will provide a brief explanation of these four selected AIs.

【🤖Nano Banana 】
Nano Banana 2(Google): An image generation AI developed by Google. It is characterized by its beautiful, realistic, and high-quality finish with detailed depictions. It is perfect for use as a base image for corrections and processing, making it easy to add touches later. If you just want something high-quality, this is the one!

【🤖Stable Diffusion】
A long-standing AI that has spread worldwide as open source. Its strength lies in its high degree of freedom and ability to support a wide range of styles. It easily produces a rugged and artistic atmosphere, and is popular among those who like its unique texture. Recommended for those who like 'addictive art styles'.

【🤖Dall-E】
Dall-E 3: Developed by OpenAI, the creators of ChatGPT. It has high fidelity to prompts and accurately reproduces what you instruct. It excels at pop and colorful finishes, with a friendly art style. Perfect for those who want it to 'make exactly what I said!'

【🤖Grok Imagine】
Developed by xAI, led by Elon Musk. Its biggest feature is that it produces unique and unexpected expressions🤣 It has a unique worldview and atmosphere that is different from other models. Recommended for those who prioritize 'surprises!'

I tried to choose ones that seem famous, but have you seen them before?

Now, let's take a look at the Pollo.ai screen operations and prompts.

👑Operation Screen, Generation Settings

Please take a look at this screen.

First, until you get used to it, please only look at the frames enclosed in red and yellow below.

🟥Red Frame: 'Headings'

You can think of this as a major 'genre classification'.
Once you get used to it, try touching it; there are various functions.
But the important part is here.

🟨Yellow Frame: 'Generation' Settings

This is the most important place.

If you create multiple times, it will become a screen like the one below.

This screen allows you to view past generations in order as you continue to generate.

However, I believe there is the same settings screen at the very bottom of this screen, so I will explain how to use this one first (*´ω｀). The content remains the same.

Here it is. Let's look at it from the top.

🟨Top: 'AI Video', 'AI Image', 'Video Agent'

This is a very rough classification of 'what to create'.
Here, we will select 'AI Image'.

🟥🟨Middle: Prompt

🟥The red frame allows you to reference an 'image'.

😯What does 'reference' mean??
For example, you put an image of a white cat in the red frame.
If you then enter a prompt like 'Make an image of the white cat from the image wearing sunglasses and riding a cool red car,' a new image will be generated that 'uses' the cat from the image exactly as prompted.

We often do this for 'video generation,' so I will introduce the details in 'Part 2' (*´ω｀)

🟥🟨Bottom: Detailed Settings

🟥The red frame only has two choices: 'AI Image' or 'AI Video'.
Of course, we will select 'AI Image'.

🟨Yellow frame
'Google Nano Banana 2' is listed here.
Yes, this is where you choose the 'image generation AI' to use.

'□3:2'
This is the 'aspect ratio'.
It refers to the screen size.
Have you ever heard or seen '16:9' or '9:16'?
These are common sizes used for portrait and landscape orientations.

'1K'
This is the image resolution.
It used to be 540P, 780P, 1080P... but now we have reached the 'K' in '1K'.

'1 on the image icon'
This is the number of images to generate.
Depending on the image generation AI, some generate 4 images at a time by default, while others generate one by one; there is a lot of variety.
Pollo.ai allows you to choose the number of generations.

What do you think?
I think it's a simple design (*´ω｀).
Everything is completed within this single generation settings frame.

④ Compose a prompt and actually try generating an image🎨

※English is recommended for AI image generation prompts.
You can generate with Japanese, but English is overwhelmingly more accurate.
To those of you thinking 'English is too much...', don't worry.
You can just have an AI translate it for you now; that's what I do too✨
Ah, but you don't have to force yourself to use English; there are those who prefer Japanese, so that's fine too (*´ω｀)✨

Now then, what kind of 'character' should I make? (*´ω｀)

Natsume asked an AI (ChatGPT, Gemini, Claude, etc.) something like this.

Well, the AI burst out laughing!
But the AI won't deny you.

I guess it is terrible after all... (*´ω｀)💦
Hey... you don't have to laugh that much... (*´ω｀)💦 Shocked

No, but I think it's a difference in perspective.
Since it's 'dark fantasy', the world is (*´Д｀)
You're really pushing the visuals at the end... (*´ω｀) Oh, I see...

So, I asked for:
・Old antique style
・Apocalyptic world

I asked for that.
Then, it created a prompt like this.

[Common Prompt]
I'll try entering the same prompt into four AIs.

'Character' settings are:

・Cute two-headed robot
・Name is 'GOMI' ← Honestly, this is unnecessary. But sometimes it gets turned into text within the image, which can be funny!
・Dark fantasy
・Old antique body
・Apocalyptic world

With just this, it builds a prompt like this.

Looking at the Japanese translation... oh... (*´ω｀)

'Eerie twilight atmosphere'... What is that? Lol 🤣

I also ask things like this, have it make corrections, and refine the prompt.

These are just collections of words, but since they can recognize sentences these days, I change how I write the prompt depending on which AI I use.

This time, I'll call it a 'common prompt', so I've prepared it in a standard AI prompt format.

If you read this, you'll see that what kind of child GOMI is and what he looks like are already entered, so let's try generating it as is (*´ω｀)

🤖Nano Banana 2

Settings complete, image size set to '3:4'. Resolution is '1K'.

And then...

It shows a screen like this saying 'Generating!'.

Oh my! (*´Д｀)
'GOMI' is inside the body (´;ω;｀)✨
And is that a sidekick riding on its shoulder!?
What a... moody dark... (´;ω;｀)

Ah, so that's what '2-head' proportions meant? (*´Д｀)

🤖Stable Diffusion 3

Settings complete! Image size... there was no '3:4', so I set it to '4:5'.

Oh my, this one feels even more 'lonely' than the Nano Banana 2 GOMI-kun...
Wait? What is this... (*´ω｀)
I feel a strong sense of kinship...
I actually felt jealous of the Nano Banana 2 GOMI-kun...

🤖Dall-E 3

Settings complete! Image size and resolution could not be selected.
I assume it will generate a 1:1 ratio.
Also, I'm not worried since I recall Dall-E 3 having high resolution by default.

Oh my, oh my, how cute! (*´Д｀)
I like big heads, you know?
Well, I guess the red lamp is a 'cheek'!
This is quite unique!

🤖Grok Imagine

Settings complete! It seems resolution cannot be selected.

Oh my! This is another expression that really captures 'despair' and a 'post-apocalyptic world'!
It's the scariest one!! (*´Д｀)✨ Yeah, yeah!
Ah... 'two-head tall'... two heads...
...Wait a minute (*´Д｀)

And so, Natsume thought.

...Aren't these all three-head tall...? (*´Д｀)

⑤ Finally

How was it?

Well, we've looked at GOMI-kun across four models (*´ω｀)

Even with the same prompt, the personalities are so different.
This is the fun part of AI image generation✨

...By the way, did you all notice?

GOMI-kun, they're all... three-head tall🤣

'I specified two-head tall!!'

Yes, this is also a common AI occurrence.
It's quite common for things not to go exactly as instructed✨
I think I can fix this by changing the 'wording' of the prompt, so it was a good learning experience for me too (*´ω｀)

But this isn't a bad thing; 'enjoying the unexpected results' is also the real thrill of AI image generation.
(Phew, glad there was a punchline (*´ω｀))

Actually, GOMI-kun might be cuter and more charming as a three-head tall character (*´Д｀)

First, try various things and generate the 'character' you want to create!
Once you feel attached to it, you're unstoppable✨

A new partner for your Note articles is born!!
You can do things like that too!! (*´ω｀)

Now, I have something I'd like to share with you (*´ω｀)

Since I focus mainly on AI image generation, many of the followers I respect and love are skilled at it, and there are many who create 'original characters' to write wonderful articles and engage in creative activities.

The articles are truly captivating, and they are a feast for the eyes (இ௰இ)✨

I am not an expert on 'character creation' or the subsequent 'management' of them, so if you have the chance, I hope you will use those creators as a reference for how they create their work (*´ω｀)✨

Huh? So, Natsume-san, what are you making now? (*´Д｀)

In a word, it's a "dark worldview."
So, it's not about "character consistency," then.
It's image generation based on a "theme" rather than a "character."

Getting to this point was the result of first getting used to operating and using AI, and then constantly facing what I really wanted to draw, which is how I arrived at "now."

As I mentioned before, at first, it was just "a woman and a cat."

Just by adding a single image to your note, your "image" changes.
Nowadays, you can generate AI images for free, so I really want you to give it a try.

Which GOMI-kun did you fall in love with? (*´Д｀)💘Just kidding

Now, we can't forget "Pollo Image 2.0", so I had it generate GOMI-kun too!!

Right: "Hey, you, your head is sinking in too much."
Left: "You too."

Natsume: "Your neck is too thin."

Pollo Image 2.0 is also quite capable of carrying a post-apocalyptic world, isn't it? 👀✨

Pollo.ai even lets you compare various AIs! ✨
Let's make your own "creations" even more fun with AI!!

That's all for this time ✨

I'd be happy if you found this interesting or enjoyed it!
Thank you so much for always liking, commenting, and adding to your magazines 🥰
Please continue to support me in the future 💐

🔽 Pollo.ai works I've created

Please feel free to read on~ (*´ω｀)✨

Well then, have a great AI image life 💐(*´Д｀)✨
Bye!! 💖

※ This article is co-written with Pollo.ai. It includes PR.

🔽 There is also a Pollo.ai mobile app ✨

iOS App

Please try downloading and using the app as well🥰✨

#Polloai #AIGeneratedImages #AIArt #ImageGenerationAI #ForBeginners #AIBeginner #AICreator #DarkFantasy #CharacterCreation #Prompt #NanoBanana #Midjourney #StableDiffusion #GrokImagine #NoteBeginner #SupportCreators