If you're interested in creating a storybook of your own and struggling with creating consistent characters, then this video is definitely for you.

如果你有興趣創作一本屬於自己的故事書，併為如何創造連貫的角色而苦惱，那麼這段視頻絕對適合你。

So to get started, you want to hop onto ChatGPT.

是以，要開始學習，您需要登錄 ChatGPT。

If you already have an account, you can click Login.

如果您已經有一個賬戶，可以點擊登錄。

If not, you'll need to sign up for an account.

如果沒有，您需要註冊一個賬戶。

So you would want to upgrade to ChatGPT 4, which will cost around $20 per month, but it is a really, really good option if you're serious about creating a storybook using consistent characters because this is one of the easiest, most uncomplicated ways I've seen where we can create very consistent characters.

是以，你需要升級到 ChatGPT 4，每月費用約為 20 美元，但如果你想使用一致的角色創建故事書，它確實是一個非常非常好的選擇，因為這是我見過的最簡單、最容易創建一致角色的方法之一。

Once you have ChatGPT 4 up and running, what you want to do is go to Explore GPTs on the side panel right here.

啟動並運行 ChatGPT 4 後，你要做的就是進入側邊面板的 "探索 GPT"。

Click on that, and we want to search public GPTs.

點擊它，我們要搜索公共 GPT。

So the one we're trying to search for is Consistent Character GPT, which is this one right here.

是以，我們要搜索的是一致字元 GPT，也就是這裡的這個。

And we're going to click on that, and now we're ready to go to start creating our consistent characters.

我們點擊一下，現在就可以開始創建一致的字元了。

To get started, you want to click on the button down here, which says Click Here to Start Creating Character Designs.

要想開始，你需要點擊下面的按鈕，上面寫著 "單擊此處開始創建角色設計"。

Once we start clicking on that, ChatGPT is going to ask us a series of questions to help us build out the character that we want.

一旦我們開始點擊，ChatGPT 就會向我們提出一系列問題，幫助我們創建想要的角色。

So the first question here is whether our character is going to be a male or female.

是以，第一個問題是，我們的角色是男性還是女性。

In our case, I think I'm going to create a female.

就我們的情況而言，我想我要創造一個女性。

Now it's asking us for a name for this character, so I'm going to say Jenny.

現在它要求我們為這個角色取個名字，所以我就叫珍妮。

And now the chatbot is asking us to share some specifics about Jenny that would include age, the country of origin, her hairstyle, outfit, and any other specific characteristics that we want this character to have.

現在，哈拉機器人要求我們分享珍妮的一些具體資訊，包括年齡、原籍國、髮型、服裝以及我們希望這個角色具備的任何其他具體特徵。

So in my case, I want Jenny to be five years old, cute little girl from Canada with brown shoulder length hair and bangs, wearing big pink headphones, plain red hoodie, black pants, and brown ankle height boots.

是以，在我的案例中，我希望珍妮是一個五歲的可愛小女孩，來自加拿大，棕色齊肩長髮，留著齊劉海，戴著粉紅色大耳機，穿著純紅色連帽衫、黑色褲子和棕色及踝高筒靴。

So let's see.

讓我們看看

So now it's asking us what type of image we want them to generate in terms of the style of image.

是以，現在它在詢問我們希望它們生成哪種類型的影像。

I would say for a children's storybook, I would either go with Pixar animation or the 2D flat illustration.

如果是兒童故事書，我會選擇皮克斯動畫或二維平面插圖。

In this case, let's try out Pixar animation.

在這種情況下，讓我們試試皮克斯動畫。

And so what should I vary in each image, whether it's just the expression, the action, which is what they're doing?

是以，我應該在每張圖片中改變什麼，是表情還是動作，也就是他們正在做的事情？

I think in my case, it would be both of the above.

我想，就我而言，上述兩種情況都會發生。

So I would say expression and action as well.

是以，我認為表達和行動也是如此。

So what expression or action would you like for Jenny in the first image?

那麼，你希望第一幅圖中的珍妮有什麼樣的表情或動作呢？

So I'm going to say, let's keep it simple for the first image.

所以我想說，第一張圖片還是簡單點吧。

I'll have Jenny standing up, looking forward with a big happy smile.

我會讓珍妮站起來，帶著燦爛的笑容看著前方。

This looks great.

這看起來很棒。

Okay, so there we have Jenny right there.

好了，珍妮就在這裡。

Now we're going to try to get Jenny to do something different.

現在，我們要讓珍妮做一些不同的事情。

Now what different variation of variables should we apply in the next image of Jenny while keeping her character consistent?

現在，我們應該在珍妮的下一個形象中應用哪些不同的變量，同時保持她性格的一致性？

We will have her sitting down and reading a book.

我們會讓她坐下來看書。

Let's try that.

讓我們試試看。

That is really good.

真不錯。

So you'll see that it is still pretty consistent.

所以，你會發現它還是相當一致的。

The only thing I noticed is that the eye color is a little bit different from the above, but overall, like the clothing and everything and the headphones, they're very consistent and you can tell that it is pretty much the same character.

我唯一注意到的是，眼睛的顏色與上面的有點不同，但總體來說，像衣服、所有東西和耳機，它們都非常一致，你可以看出這幾乎是同一個角色。

And now what we can do is let's have her running with her hands up in the air with a very happy and excited expression.

現在我們可以做的是，讓她高舉雙手，帶著非常高興和興奮的表情奔跑。

Let's try this.

讓我們試試這個。

This looks great.

這看起來很棒。

And now let's try a fourth one.

現在我們來試試第四個。

What do we want Jenny to be doing?

我們想讓珍妮做什麼？

Let's try something harder and see if they're able to generate that for us.

讓我們試試更難的，看看他們能否為我們創造出這樣的機會。

Let's have Jenny riding on a pony.

讓珍妮騎上小馬。

Let's see if that works.

讓我們看看這是否有效。

And that is pretty good.

這已經很不錯了。

I think the one mistake I made was that I didn't specify the eye color.

我想我犯的一個錯誤就是沒有說明眼睛的顏色。

And I think I could have probably been more descriptive with my boots, whether they were laces or not.

我想我本可以更詳細地描述我的靴子，不管它們有沒有鞋帶。

And I think that would basically help eliminate the issues that we see.

我認為這基本上有助於消除我們所看到的問題。

And there are very minor issues between some of these photos where this version had brown eyes and this one has more of a greenish color and in the boots.

其中一些照片之間還存在一些很小的問題，這個版本的眼睛是棕色的，而這個版本的眼睛和靴子的顏色更偏向於綠色。

These ones have laces.

這雙鞋有鞋帶。

I think most of them have laces with the exception of this one.

我想除了這雙鞋之外，大多數鞋都有鞋帶。

So I think by incorporating that into my original prompt here, when they were asking us to provide some details about Jenny's features, I could have probably added brown eyes and also brown ankle height boots with laces and that would have eliminated these inconsistencies.

是以，我認為在我最初的提示中加入這一點，當他們要求我們提供一些關於珍妮特徵的細節時，我也許可以加上棕色的眼睛和棕色的帶鞋帶的及踝靴，這樣就可以消除這些不一致的地方。

But overall, I think this is actually really, really amazing.

但總的來說，我覺得這真的非常非常棒。

If you look at it, they all look like consistent characters just doing very different things.

如果你仔細觀察，它們看起來都是一致的角色，只是做的事情截然不同。

And it is so easy to create.

而且製作起來非常容易。

There was no need to do anything with seed numbers or upload a bunch of other images.

無需對種子號碼做任何處理，也無需上傳大量其他圖片。

It is really just answering the questions that are prompted by ChatGPT and they automatically created these images.

其實就是回答 ChatGPT 提示的問題，他們就會自動生成這些圖片。

So now what you would want to do is to download all the images that we had created.

現在你要做的就是下載我們創建的所有影像。

In this case, we had created four, but you can create more if you'd like.

在這種情況下，我們創建了四個，但如果你願意，也可以創建更多。

And the other thing I should mention is when ChatGPT creates the images, they usually create four.

還有一點我得提一下，ChatGPT 製作圖片時，通常會製作四張。

And then once they created four, they'll ask you to override some of the previous ones if you create any more.

然後，一旦他們創建了四個，如果你再創建其他的，他們就會要求你覆蓋之前的一些。

So what I would usually do is just download the previous ones that I had created and then just allow them to override them after.

是以，我通常的做法是下載之前創建的程序，然後讓他們覆蓋這些程序。

So once we've downloaded all the images, the next step you want to take is to upscale these images just because you want them to be at the best quality possible before you use them for your storybooks.

下載完所有圖片後，下一步要做的就是放大這些圖片，因為在將它們用於故事書之前，你希望它們的品質儘可能好。

So the tool I'm going to be using today is Upscales.

我今天要使用的工具是 Upscales。

It's a free tool to allow you to upscale the images.

這是一款免費工具，可讓您放大影像。

Let's try uploading one of my images.

讓我們試著上傳我的一張圖片。

So you can choose two times or four times.

是以，你可以選擇兩次或四次。

In my case, I'm just going to do two times and we're going to click download.

在我的情況下，我只需要做兩次，然後點擊下載。

Let's see.

讓我們看看

Okay, so that's done.

好了，就這樣吧。

And you basically do the same thing for all of your images.

基本上，你要對所有圖片做同樣的處理。

So once you've upscaled all of the images, they're now ready for us to use in our storybook.

是以，一旦你放大了所有圖片，它們就可以用於我們的故事書了。

And in the next step, I'm going to show you how to incorporate them into a storybook with a different background using Canva.

下一步，我將向你展示如何使用 Canva 將它們融入一本不同背景的故事書中。

So once you get to Canva, the first thing you want to do is upload the images that we had just created.

進入 Canva 後，你要做的第一件事就是上傳我們剛剛創建的圖片。

As you can see, I have already uploaded the four on my end.

如您所見，我已經上傳了我這邊的四張照片。

But if you haven't already done so, you can just click on upload files and upload your images.

但如果您還沒有這樣做，您可以直接點擊上傳文件，然後上傳您的圖片。

And so for illustration purposes, let's say I want to use this one to create a scene for the story.

是以，為了便於說明，假設我想用這幅畫來創作一個故事場景。

Within Canva, obviously, you can pick a background.

顯然，在 Canva 中，您可以選擇背景。

You go to elements and let's say, you know, I want a library background.

你去找元素，比方說，我想要一個圖書館背景。

Let's say I could use this one.

比方說，我可以用這個。

Or alternatively, you could also create a background from Mid Journeys or any other AI generator.

或者，您也可以從《中途旅程》或任何其他人工智能生成器中創建一個背景。

So you could describe the background that you want and create it that way.

是以，您可以描述您想要的背景，然後創建它。

I've actually picked a few backgrounds that I'm going to use for illustration purposes already.

實際上，我已經選好了一些背景，準備用來做插圖。

So I'm actually going to use those.

所以，我真的打算用這些。

But just wanted to say that there are multiple options for you to use from a background's perspective.

但我想說的是，從背景的角度來看，您可以有多種選擇。

So here I have already created a background image that I want to use for the scene that I'm creating.

在這裡，我已經創建了一張背景圖片，我想把它用於我正在創建的場景中。

So let's hop on back to uploads where I can find my images and I want to pick this one like she's reading in the library.

所以，讓我們跳回到上傳，在那裡我可以找到我的圖片，我想選這一張，就像她在圖書館閱讀一樣。

I'm going to click on edit photo because first you need to remove the background.

我要點擊編輯照片，因為首先你需要移除背景。

This is really simple within Canva.

這在 Canva 中非常簡單。

All you need to do is go to background remover, click on it and it automatically removes the background for you flawlessly.

你只需進入背景移除程序，點擊它，它就會自動為你完美地移除背景。

So now all you have to do is adjust it where you want to position.

現在你要做的就是把它調整到你想要的位置。

The girl can really be anywhere you want.

女孩真的可以在任何你想去的地方。

Just make sure it's proportionate to the background.

只要確保與背景相稱即可。

And then let's say, you know, we want to create a second one.

然後比方說，我們想創建第二個。

This one's in the park.

這個在公園裡。

So I think this one would be quite suitable for kind of running across the park.

所以，我覺得這個很適合在公園裡跑來跑去。

Similarly, we go to edit photos, background remover, and there we go.

同樣，我們去編輯照片，去掉背景，就可以了。

Again, we want to make sure it's proportionate.

同樣，我們要確保它是相稱的。

There we go.

好了

And then I can add text and we can adjust the size.

然後我就可以添加文字，調整文字大小了。

So let's say Jenny runs across the park excitedly to see her friends.

比方說，珍妮興奮地跑過公園去看朋友。

And that could be one scene within your storybook.

這可能就是你故事書中的一個場景。

Change it to, let's say, white.

比方說，換成白色。

I'm going to move it to either the bottom or the top, align it differently.

我要把它移到底部或頂部，以不同的方式對齊。

And there we go.

我們走吧。

That's one page.

這是一頁。

And you can also add other things within Canva, right?

你還可以在 Canva 中添加其他內容，對嗎？

If you want her to have, you know, a dog, you can also add that.

如果你想讓她養狗，你知道，你也可以加進去。

You just want to make sure that the style of the animation is consistent with the animation you have here.

你只需確保動畫的風格與這裡的動畫保持一致。

It doesn't really suit, but, you know, just as an example, you can add other things to the scene.

這其實並不適合，但是，你知道，舉個例子，你可以在場景中添加其他東西。

So it might take some searching around or alternatively, you could use the same chat GPT program we used earlier to create the dog image as well.

是以，您可能需要四處搜索，或者也可以使用我們之前使用的哈拉 GPT 程序來創建狗狗影像。

Then that way you'll make sure it's kind of like that Pixar animation style.

這樣你就能確保它有點像皮克斯的動畫風格。

And that's really it.

就這樣。

I hope you enjoyed this tutorial and you found it useful.

希望您喜歡本教程並覺得有用。

And I can't wait to hear about all the storybooks that you will be creating.

我迫不及待地想知道你們將要創作的故事書。

If you enjoyed this video and found it helpful, please don't forget to subscribe so you won't miss out on my other videos.

如果您喜歡這段視頻並認為對您有幫助，請不要忘記訂閱，這樣您就不會錯過我的其他視頻了。

I'll see you next time.

下次見