Online AI Courses for Kids-Use generative AI to Draw Paw Patrol Dog Chase
Hi parents and kids, this free online AI course "Use generative AI to Draw Paw Patrol Dog Chase" will gives your K12 kids a brief introduction to what is Artificial Intelligence (AI) technology, what can AI accomplish (such as AI Generated Contents converting text to image, Chatbots giving automatic response), and how does AI work in the easiest way that even K12 children can understand. This course covers some fundamental concepts of AI, including what neural networks is, what's the most popular neural network blocks of AI models (Transformers, Diffusion, etc.). Finally, kids can also participate by interacting with the AI during the course, by clicking the image or text with 'robot' icon or button, and get more detailed explanation, type a query or ask question using microphone by themselves, and accomplish the task of generating a picture of unique Paw Patrol Dogs and get the trophy. This course will takes around 20 minutes to complete. And all the interactive learning sessions in the course are designed by experienced children-education professionals. Let's get started!
Instructions of Interactive Sessions
- Click on the text or image with the robot icon for detailed term explanation
- Click the media player icon for automatic reading of the course for your kids
- Click the button to get AI generated language replies or images
Keywords: AI Courses,K12,Age 6-9,Age 10-13,AIGC,Text to Image,Paw Patrol
SYLLABUS
- 1. What is Artificial Intelligence
- 2. What can AI accomplish
- 3. How does AI generate a picture
- 4. Let's Try: Ask AI to Draw Picture of Paw Patrol Dogs Chase!
-
1. What is Artificial Intelligence
Hello kids, I am your course instructor and my name is Kate. Today I will teach you a lesson about what is Artificial Intelligence. Artificial Intelligence (AI) is the technology to use computers to imitate human's behaviors as if they have real intelligence as human being. AI has many capabilities including: Language, Vision, Reasoning, and Memory, etc.
-
Language
Language or NLP(natural language processing) is a subfield of artificial intelligence. There are many implementations such as Part-of-speech tagging, Named Entity Recognition, Machine translation, Automatic summarization, etc. See NLP for more details.
Language ability means AI can talk and reply in natural language just as humans. -
Vision
Vision ability refers to AI can analyze a picture and describe what's on the picture just like humans. -
Reasoning
Reasoning
ability refers to the situation that AI can do complex math, e.g. add, substract, multiply and more advanced calculations. -
Memory
Memory
Memory ability means that AI can memorize a lot of knowledge, even the encyclopedia, in the computers' memory by compressing the knowledge into bits and keep it in computers' storage.
-
-
2. What can AI accomplish
There are many things you can ask Artificial Intelligence to help you finish. Doing your homework? Probably. But scientists invent AI for many other interesting things. Here is a few applications of AI in real world scenarios.
-
Chatbot or AI Assistant
You can use chatbot in the way of replying to text message, and ask questions of local weather, make booking of flight tickets and take-out food, etc.
Image Generation (Text to Image)
You can generate a image by simply making detailed descriptions about what you like. For example, if you want to draw a picture of your favourite Paw Patrol Dog, you can ask AI a demand "Draw a picture of Paw Patrol dog Chase with a police car". And you can get your unique picture.-
Video Generation (Text to Video)
You can even generate a short CLIP of video by making description verbally. OpenAI company just released their latest AI model SORA, which can generate a video of up to 1 minute length.
-
-
3. How does AI generate a picture
The whole process of AI generating a picture is really complicated. But it consists several common steps that you can understand.
- First of all, the AI will needs to convert all the image or text into vectors of numbers or embeddings so that AI can understand as inputs.
- Secondly, AI models use neural networks to convert the embeddings of text and image and make calculation of many probability, like the probability of the next word that human is going to say, and the next small patch of pixels of image that human is goind to focus on.
- Finally, the output probability will be used to predict the final output of the most likely sentence or the most likely pixels of images. And AI will generate the final sentence or images.
-
Transformer
Transformer models don't refer to the transformer in the movie, but refer to the most popular AI models in the Natural Language Processing and Computer Vision societies. It is invented by Google and becomes popular afterwards.
The "Transformer" Movie
Transformer Neural Network
Transformer model architecture from paper "Attention is All You Need". See this paper for more details.
-
Diffusion Models
Diffusion models refer to the type of AI models, which takes input of an image, add random noise to the image's embedding step by step. Once it reaches the hidden layer z and the model will reversely reconstruct or generate the original image.
-
4. Let's Try: Ask AI to Draw Picture of Paw Patrol Dogs Chase!
- Step 1. Draw your Favourite Dog: Chase
- Step 2. Add a Police Car
- Step 3. Add the Background: In the Park
- Step 4. Finally, add your kids' Name (e.g. Kevin) to this unique drawing
Congratulations, Kevin! This is your AI generated Paw Patrol Picture. Just Save it and Make a Greeting Card!
- Step 1. Draw your Favourite Dog: Chase
-
You can also try the chatbot service simply clicking the button below this blog. Wechat users can also follow the wechat public account "深语人工智能" after scanning the QR code