Lesson 6.3: Explore Nous by the Activities

1. Lesson Objectives: Further explore Nous's features and applications through eight activities, including Nous's pre-set AI functions, graphical display functions of Nous Screen, Tiny ML (speech model & image model), Autopilot, Color Tracking, and Visual Line Following..

2. Time: 90 mins

3. Language: English

4. Format: Completely Online

5. Level: Advanced

Unit List:

1. MatataChat & MatataDraw

2. Experience basic computer vision in Nous

3. Nous loves to Draw

4. Nous Pet

5. Nous Knows Animals

6. AprilTag-Based Autopilot

7. Color Tracking

8. Visual Line Following

Unit 1. MatataChat & MatataDraw

The Nous HUB provides a series of interesting default programs (e.g. MatataChat and MatataDraw) you can easily experience with one simple click.

Before using MatataChat or MatataDraw in a Nous HUB for the first time, you need to get an authorization code from your connected PC/tablet.

1. Connect your Nous HUB to MatataCode.

2. In the top menu, find “AI” and select “Authorization code”.

3. Click the authorization icon to complete the authorization.

1.1 How do I use MatataChat?

1. Press the MatataChat icon on your Nous HUB, select Wi-Fi, and then enter the Wi-Fi password (if needed).

2. After successful WiFi connection, press “Chat” and ask your Nous HUB questions as prompted.

3. Wait for about 20 seconds, and the Nous HUB will answer your questions by voice.

Note: You should first select a language for speech recognition and response in the top-right corner.

Task: Students will experience the functions of MatataChat and ask “What date is today?” in their native language.

1.2 How do I use MatataDraw?

1.Press the MatataDraw icon on your Nous HUB, select Wi-Fi, and then enter the Wi-Fi password (if needed).

2.After successful Wi-Fi connection, click the “Input” button, and describe the picture you want.

3.Click “Draw”.

4.Wait about 20 seconds, and the Nous HUB will display the picture on-screen.

Task: Students will experience the functions of MatataDraw. Describe the picture named “cute cat”, and press “draw” to get the picture.

Unit 2. Experience basic computer vision in Nous

2.1 Human face recogniton

Task: When the program is running, turn on human face detection. If a face is detected, Nous says "Hello, I am Nous!", displays"Hello" on the screen, sings "twinkle,twinkle,little star", and Nous walks forward for one second and walks backward for one second at 50% speed.

2.2 AprilTag recognition

Task: When the program is running, AprilTag detection is turned on. When Tag1 is detected, Nous plays the sound effects of an ambulance until done; when Tag2 is detected, Nous plays the sound effects of a police car until done; when Tag3 is detected, Nous plays the sound effects of a fire engine until done; When Tag4 is detected, Nous plays the sound effect of a train until done.

2.3 MNIST handwritten digit 0-9 recogniton

Task: When the program is running, turn on the MNIST (handwritten digital 0-9 ) detection. When Nous detects the number 1, it says "1" and displays the number 1 on the screen; when Nous detects the number 2, it says "2" and displays the number 2 on the screen...

2.4 Color recognition

Task: When the "A" button is pressed,set color block categore to red(L:20,80,A:30,50,B:15,35), and color detection is turned on. When a red card is detected, a red circle is displayed in the center of the screen.

Activity Two: Experience basic computer vision in Nous

Unit 3. Nous loves to Draw

The screen doesn't just display characters, variable information, etc., but also displays patterns.

Before displaying something on the screen, we need to first understand how the screen displays patterns. Here, we must mention two important concepts: pixels and resolution.

Pixel: simply stated means a point, many pixels can be combined to form the image we see.

Resolution ratio: Resolution ratio is used to indicate how many pixels there are horizontally and vertically

The resolution ratio of the Nous screen is 320x240. Based on this information and the knowledge of coordinates, we can determine how to display some basic patterns on the screen. For example, use the center point of the screen as the circle point to draw a circle with a radius of 40. Then, the coordinate information of this center point is: X=160, Y=120.

The concentric circles with no fill color

The concentric circles with filled color

The Olympic Rings

Activity Three: Nous loves to Draw

Unit 4. Nous Pet

4.1 What is Tiny ML?

Get to know the four steps of Tiny ML

The 4 steps of Tiny ML：Through Model Creation，Data Acquisition, Training&Development and Programming, Nous can recognize Images,Speech,etc. , while no Internet or PC required!

Nous’s 3 typical Tiny ML models: Image Model,Speech Model ,Tof Model.

4.2 How to Create a Speech Model?

Step 1: How to choose and create a speech model?

Step 2: How to acquire the data?

Design the commands/wake-up words for the Nous Pet

Before acquiring the audio data, make sure the Nous Robot is connected to the coding platform.

Change "Class 1" to "Hi Nous" and check the "Setting".

The "delay" refers to the delay of several seconds to start collecting data after pressing the "Acquire" button.

The "duration" refers to the duration of time to acquire one sample. Usually, we need to repeat the commands/wake-up words several times in this duration.

In this case, we will use the default settings: the delay is 0 second, while the duration is 10 seconds.

Press the "Acquire" button to acquire the data samples. There are several tips:

1. During the "duration" of acquiring the data sample as the duration in this case is 10 seconds, you need to repeat the command/wake-up word repeatedly;

2. Data diversity: the more diverse the data, the better the model. Thus, we can acquire speech data with different speaking speeds, timbres, or accents;

3. Please collect at least 4 audio data samples for each class.

Add 4 more classes, including "Go forward", "Go backward", "Spin" and "Nothing", and acquire at least 4 data samples for each class, just like you did for the first class "Hi Nous" . Remember, "Noth_x0002_ing" means there is just background noise, so we just need to press the "Acquire" button and acquire the background noise.

Step 3: How to train and deploy the data?

How to program using the Speech Model?

Activity Four: Nous Pet

Unit 5. Nous Knows Animals

Note：Before acquiring the data, make sure the Nous Robot is connected to the coding platform. What's more, please connect Nous to the same wireless network as this device (PC/Pad).