Ancient poetry this play in fact many people, have seen on the Internet, there are also many students have learned by themselves, then want to innovate, and how to get out of their own personalized play? Here in the teaching students, thought of combiningDigital HumanThe playfulness of the game, using images of characters and authors talking to each other to create a different way to play with the learning of ancient poems.
Okay, here's where we get into the practicalities, the content is dry, so I suggest you make sure you save it.
There are AI tools that need to be used on this:
1. Raw map tool: i.e. Dream
2. Video tools: i.e. Dream
3. Editing tools: Cutting and mapping
4. Cue word writing: DeepSeek, beanbag
Step 1: Create an ancient poetry selection + cue words for each subplot
The first situation: here we do not know what to do in the case can be directly thrown to DeepSeek to recommend, remember the universal formula:What are you going to do + what is the purpose + what do you want to achieve + what do you want to keep him from showing up.
For example, the following one: I want to make an AI ancient poetry video for 0-6 years old baby's ancient poetry enlightenment, but I don't know which are for 0-6 years old ancient poems, please help me recommend 10 ancient poems. Or only he help me filter out 10 popular ancient poems for primary and secondary school students
With the above ancient poem in place, we then dialogued to get him to come up with the split-screen cue word, and here, for testing purposes, we purposely used the Qingming ancient poem.
Help me generate the AI painting cues for the first "Qingming" ancient poem split screen. Each line of the poem should correspond to a scene. Include the main scene, style, quality words, each prompt word last add "ancient Chinese cartoon, bright colors, 3d Pixar style", the prompt word should be simple and easy to read, do not contain complex descriptions, do not have adjectives and lyrical part, do not appear the name of the person, each screen need to appear characters, always a girl, wearing a pink Tang suit clothing, please describe the appearance of the character (fixed). Pink Tang suit costume, please describe the character's appearance (fixed), clothing (fixed), and behavior. Then according to the descriptor to generate the picture to video cue word, cue word on the logic according to the subject what to do. Output in table form.
Wait a little while for the AI drawing prompts for each verse of Qingming and the prompts for the Tusheng video to appear!
The second situation: is directly know what we want to do what ancient poetry: the Qingming this ancient poem, give the original ancient poetry and each sentence corresponding to the modern translation, sent to the DK, then we are in the request: please help me according to the content of the above Qingming poems, each line of the ancient poetry, help me to generate a dream AI sub-scene prompt words, the requirements of each screen is to be the ancient Chinese cartoon style, bright colors, 3D Pixar style, Q version! Each line of the poem should be matched with a scene and a child character. This way, we will also get the split-screen cue words for each verse.
The third situation: the use of AI intelligent body, here we do not teach you how to build their own, we directly use others to do on the line here to cheer the duck teacher went to search for the beanbag, found that there are really quite a lot.
Open Doubao - More on the left - Select Smart Body - Enter the ancient poem prompt word search
Here choose 3D little girl test. Input the ancient poem to be done, the intelligent body will give us the poem screen text born picture cue words, picture born video cue words.
Through these previous 3 ways we can quickly get the ancient poem prompt words we want to make. Completely without hands to write, zero basis can also come up with words.
But today we don't do poetry alone, we also innovateDigital Human DialogThe two figures will also need to be configured.
Character Cue Words: a girl with a double bun, fair skin, big eyes, wearing a pink Tang costume outfit, standing in front of a wooden table, looking at the camera, ancient Chinese cartoon, 3d Pixar style, solid color background (go ahead and rewrite the Save Character Consistency based on the generated cue words)
Author's cue: image of the ancient poet Du Mu, front view, solid color background
Of course if you can't write it, you can have the AI write it:
Kid: "Mr. Du Mu! In your poem, you said, 'It rains a lot during the Qingming Festival', is it so accurate every year?"
Du Mu: "The God of Heaven knows that this rain...is just like the undried tears on earth."
Kid: "We always bring flowers and green dumplings to the graves, did you do the same in your time?"
Du Mu: "Fire is forbidden during the cold food, only cold porridge is offered to the ancestors. Do you know that it is not the rain that breaks the souls of those who walk on the road, but the fact that I don't see the people of the olden days?
Kid: "The 'Apricot Blossom Village' you're looking for! I've checked to locate a real wine shop thirty miles away!"
Du Mu: "I don't want to think that after a thousand years, a childish boy will become a shepherd boy pointing the way, but if I can keep my poetic heart, I will be able to find apricot blossom villages everywhere on earth."
This way the cue words in the video are fully organized.
Step 2: Cue word AI tool out of the picture
Can generate AI pictures of the tool there are many, but also to you to organize can be sent, here or with that dream demonstration. Address: https://jimeng.jianying.com
Raw Chart Model Selection 3.0
Input prompt words: girl (round face with almond eyes/double hair bun with red ribbon/pink Tang suit/pink embroidered shoes) standing on the green stone road with an oil-paper umbrella, drizzle falling, hazy mountain shadow in the distance; ancient Chinese cartoon, bright colors, 3d Pixar style, high definition image quality
Input the prompt word: girl (round face with almond eyes/double hair bun tied with red ribbon/pink Tang dress/pink embroidered shoes) walking on the wet path with her head down, the corners of her clothes stained with mud, the weeping willows on both sides swaying, blurring the back of the pedestrians in the mist; Ancient Chinese cartoon, bright colors, 3d Pixar style, high definition picture quality
Input prompt words: girl (round face with almond eyes/double hair bun tied with red ribbon/pink Tang suit/pink embroidered shoes) wet path, clothes corners stained with mud, stooping down to ask the cowherd boy riding a cow, holding a bamboo hat, the cowherd boy lifts his hand and points to the faraway place; Ancient Chinese cartoon, bright colors, 3d Pixar style, high definition picture quality
Input prompt words: girl (round face with almond eyes/double hair bun tied with red ribbon/pink Tang dress/pink embroidered shoes) turns to look into the distance, pink apricot blossom forest reveals the roof of the village, the wine flag flutters in the wind; ancient Chinese cartoon, bright colors, 3d Pixar style, high definition picture quality
So that we get their own satisfaction with the poetry map, choose their own satisfaction with the image without watermark to save. Also measured under the intelligent map effect is also very good, we use by themselves.
The most core can also give you a cover picture (very fragrant, will not be able to write can find the next cheer duck teacher to make up the XX cover descriptor)
Character part of a raw drawing demonstration:
Step 3: AI Tools Tupelo Video
Because the video action amplitude is very small, the direct use of the dream is complete, here is not overly detailed show, of course, other video tools are also available.
Ancient Poetry Section
Counterpart part (Counterpart part is recommended to have enough points to use all master level, here there are not enough points to use the vivid mode, the system will crop)
Here, we basically the entire video needs material and video clips have been completed, the next last step editing. Many people say that 30% relies on content and 70% relies on editing.
Step 4: Edit the finished product
1, open the cut screen PC Professional Edition, enter the main interface, click [+ start creating].
2, add the saved video clips, add background music, you can directly search in the cut screen
3、Appropriate location to add the right sound effects, text information, you can double the effect of the video.
See the finished product and compare the results
Cross-screen real-life live dialog effects.
Finally think the tutorial is good can give a small reward, stay up all night to code words is really not easy, the core is everyone's innovation, clips can add points.
Cautions and Operational Recommendations
1. Maintain originality: Do not directly copy the content of other people's cue words, by modifying the creative body of the cue words or adjusting the details of the screen video, to ensure that the work has a certain degree of originality.
2. Optimize the rhythm of the video: too long animation content may lead to viewers skipping, it is recommended to control the video time to ensure a compact rhythm.
3. Accurately locate the audience: choose a platform and time period where the target audience is clear to publish the video, for example, uploading it in the evening for better results.
4. Interaction to enhance fan stickiness: actively reply to fan comments, increase the interaction rate, attract more users to pay attention to and forward.