The little guy sent it.AI Wellness VideoBegging to be disassembled, AI made a wellness video that's hot as hell.
In the major video platform search "health", appeared a number of new accounts, each are hundreds of thousands of powder. For example, the following account 20 works, up 400,000 powder, video strip explosion, the highest one 430,000 likes.
The health and wellness track has a large audience, especially the middle-aged and elderly population nowadays. Using AI to produce popularized videos of Chinese health care knowledge from various canons in the Yellow Emperor's Internal Canon, such videos explode at the drop of a hat.
This track is more flexible in terms of realizing revenue: for example, window banding, creator share revenue, teaching people to make videos, and this number owner, window banding has sold tens of thousands of books.
Today, we teach you to dismantle the production method of this kind of AI permaculture video, this is a lot of dry goods, it is recommended to collect attention to prevent lost.
Step 1 DeepSeek Generate Permaculture Copy
The first step in making a video is to choose a topic and copywriting, whether the short video explodes or not, the choice of topic accounts for half.
Copywriting we can look for the counterpart pop-ups, let deepseek to mass production. Search for the selected copy of popular pop-up videos and let the AI analyze and output new pop-up copy accordingly, so that the videos produced will be easier to get out of the circle.
Open deepseek's Deep Thinking and Networking feature and enter the following prompt words:
Address: chat.deepseek.com
- Input: I want to make an AI video on TCM health, please refer to the 2 videos I provided.short video that explodes in popularityPermaculture Copywriting, help me generate 5 permaculture video copywriters
The prompt is followed by a reference text for deepseek. In principle, the more reference texts you feed to deepseek, the better, and the more explosive, the better.
DeepSeek generates us 5 selections and corresponding video copy after deep thoughts, we check them out and find the ones we are happy with to work on, and continue generating if we are not.
Based on the structure of the bomb case you provided, I sum up the creative formula for "Strike-to-life" and produce the following five video files for medical care:
1. [Comparison table for the four seasons of health care]
Rubbing the tendons in the spring is the best way to protect the liver, and drinking green bean soup in the summer is the best way to beat the heat!
Eating silver fungus in the fall is the best choice for moistening the lungs, and eating lamb in the winter is the best way to drive out the cold
Standing against the wall when you are sedentary is the best spine care, and patting your armpits when you are angry is the best liver secret.
Drinking wolfberry chrysanthemum tea after staying up all night is the best remedy, and turning the eyeballs when you use your eyes a lot is the best way to improve your eyesight
2. [Guide to Physical Conditioning]
Drinking Chenpi water for phlegm-damp body is the best way to get rid of dampness, and drinking lily porridge for yin-deficient body is the best way to nourish yin.
Yang deficiency body tanning back is the best yang tonic, qi depression body press tai chong acupoint is the best depression relief techniques
Drinking rose tea is the best way to invigorate blood stasis, and practicing Baduanjin is the best way to replenish qi.
Allergic body press foot Sanli is the best anti-allergy art, insomnia body rub Chungquan point is the best sleep know-how!
Step 2 DeepSeek Generate Drawing Cues
Next, we'll have DeepSeek generate the sub-scenes into AI painting cues that can be easily generated into the AI tool, and I'll use the seasons of health and wellness cross-reference table as a selection to create:
- Please optimize the detailed scene-by-scene script according to Midjourney syntax rules, including screen details, style parameters and camera language, and generate cues for 8 scenes, keeping the same characters, styles, and attire, and outputting in Chinese. Add "Chinese ink painting, Xu Beihong" to each cue.
DeepSeek gives painting cue words:
Below are 8 AI painting descriptors for Split Screen that match the requirements:
Image 1: Anxious woman with banana
Chinese style ink painting, borrowed from the style of Xu Beihong, an ancient scene of a 20 year old woman dressed in traditional Chinese hanbok. The woman, with a slight frown and anxiety revealed in her eyes, sits in an ancient courtyard with a bunch of golden bananas on a stone table. She gently picks up a banana and is about to peel it, when the breeze brushes across her coat, and the flowers and plants around her sway slightly. The overall color tone of the picture is on the light side to render a quiet and slightly sad atmosphere. A medium shot is used to show the interaction between the woman and the banana.
Image 2: Depressed Woman with Citrus
Chinese ink painting, borrowed from the style of Xu Beihong, of a 20 year old woman dressed in a traditional Chinese hanbok, in an ancient style scene. The woman looks forlorn, her head hanging low, standing next to a bamboo basket full of mandarin oranges. She slowly approaches the mandarin oranges and sniffs their fragrance, which seems to be slowly awakening her depressed mood. The sunlight pours on her through the gaps in the leaves, creating dappled light and shadows. Use a close-up lens to focus on the woman's facial expression and the tangerine.
Image 3: Irritated Woman with Sour Plum Soup
Chinese style ink painting, borrowed from the style of Xu Beihong, an antique scene of a 20 year old woman in traditional Chinese hanbok. The woman sits in an antique interior, her brow furrowed, looking annoyed. On the table is a steaming bowl of sour plum soup, which she picks up, blows gently on it, and then takes a small sip. The screen, paintings and other decorations in the interior add to the old-fashioned flavor. Close-ups are used to highlight the color of the plum soup and the woman's movements.
The other specter omitted..
Harmonization of parameter descriptions:
- Core Painting Style: Xu Beihong Rendering Ink and Wash (Accurate Animal Forms + Landscape Context)
- Character Set: 30 years old woman/Indigo blue rough jacket/low bun without jewelry
- Scene features: farmhouse elements (bamboo ware/stone mill/old acacia/ celadon)
- Color system: black, white and grey only + monochrome accents (anxiety yellow/depression orange/fire-reducing purple)
- Dynamic rendering: 0.5 levels of motion blur for key elements (banana peeling/steam rising)
- Taboos: eliminating electronic devices/modern clothing/highly saturated color blocks
Once the drawing cues have been generated, we next go to the AI generation tool to create them.
The third step, that dream to generate images
So let's do it with Instant Dream AI, which gives you 66-88 points per day, which is equal to free usage.
i.e. Dream AI: https://jimeng.jianying.com
- Steps:
- Open the webpage (or mobile APP) of Imagine Dream, click the image to generate
- Model selection picture 2.0pro model
- Ratio adjusted to 9:16
- Enter a split cue, where we can use an AI-generated painting cue or write our own.
- Click Generate Now.
We took each of the split-screen painting cues generated by deepseek and copied them into the input box, the
The AI generates 4 images at a time, from which we find the ones we are happy with, e.g. I think the first one is good, click on the image, click on the click on the Ultra HD and export the image.
This image serves as our reference image for generating the split screen later.
Following this approach, we took the remaining subplot descriptors and fed them all into Instant Dream AI to generate the full subplot screen.
Step 4: Image Generation Video
Next, we're still generating videos in Instant Dream AI.
Find our generated image and click on Generate Video, the page will adjust to the Figure Generated Video window.
In the video generation interface, we directly copy the AI painting prompt words into the input box.
Wait a little while and the generated video will be displayed in the page. We can check if the video meets our requirements and click Generate Again if we are not satisfied.
Repeat the previous step to input the remaining split-screen descriptors into the Instant Dream AI, first into an image, and then into a graph-generated video.
Note: If Dream AI does not open a member, the downloaded video has a watermark, we can use Tencent Smart Shadow, HitPaw Watermark Remover and other AI tools to remove the watermark.
Step 5: Generate the voiceover
The video is there, next a voice over.
Go to TTSMAKER online dubbing tool, input the ancient poems into the input area, choose Hot Recommended Generic Female Voice Dubbing, and click the bottom right corner to start the conversion (free).
TTSMAKER: https://ttsmaker.cn/
Click to download the file locally.
Step 6: Edit the video
The last step is to go to the clip and composite the video.
Import all the split-screen video clips, import the voiceover generated in step 5, and add text subtitles.
Choose a classical country style music and our video is made.
Methods of realization
This track is more flexible in terms of realizing revenue: e.g., window banding, creator share revenue, teaching people to make videos, and so on.
1.Traffic Master Realization
Video platforms all have creator share programs, where videos are earned when they are played
2The Window Strip
Opened a window to sell health books, you can see that Wang Yangming books sold 144,000 copies, and Yangmou sold 700,000 copies.