Face recognition

Operation steps

1. Tap the Face icon on the AI Camera 2.0 to start the face recognition feature.

   

2. Orient AI Camera 2.0 toward the face for recognition. For a new face, AI Camera 2.0 will frame it in white and display Unknown. The top menu shows available operations for the current face: Face Detection, Face Learning, and Emotion Recognition. You can tap the corresponding option or toggle the switch on the top left or right to switch between them.

Aim the cross in the center of the screen at the face, then press the Learn button on the top of the device or tap the learning icon.

After learning, AI Camera 2.0 marks the face with a green box and displays it as Face1. The number following the name represents the recognition confidence. In the example below, the confidence is 0.89.

Face deletion

Face detection

Expression recognition

Coding block example

Face recognition blocks

Color recognition

AI Camera 2.0 can learn and recognize colors. It can obtain information such as a color’s coordinates (X, Y), width and height (W, H), and quantity. This feature, together with the small balls included in the package, helps you create color-based sorting projects.

Operation steps

1. Start the color recognition feature:

   Method 1: Tap the Color icon on the AI Camera 2.0 screen.

2. Aim the cross in the center of the AI Camera 2.0 screen with the target color block. Press the Learn button or tap the color area. AI Camera 2.0 will automatically frame the block and display its L, A, and B values.

Color learning

Color switching

Color deletion

Coding block example

Gesture recognition

AI Camera 2.0 comes with 12 built-in gestures and recognition cards to support gesture recognition. It provides the coordinates of each key point, enabling more advanced interactive features.

In addition to gesture recognition, AI Camera 2.0 also supports gesture learning, which allows you to create fun projects such as interactive robots.

Operation steps

1. Start the gesture recognition feature:

• Method 1: Tap the Gesture icon on the AI Camera 2.0 screen.

• Method 2: In mBlock’s General category, use the Start Gesture Recognition block to start the feature.

2. Use gesture recognition features.

Gesture recognition

Gesture learning

Gesture deletion

Make the gesture you want to delete and press the Learn button or tap the forget icon. The screen will prompt Confirm to forget?. Tap Confirm to delete the gesture.

Coding block example

Tag recognition

AI Camera 2.0 can recognize AprilTags, QR codes, and barcodes. For AprilTag recognition, AI Camera 2.0 can detect both the distance and rotation angle of an AprilTag relative to the camera. With tag recognition, you can create vision-based positioning projects.

Operation steps

1. Start the tag recognition feature:

• Method 1: Tap the Tag icon on the AI Camera 2.0 screen.

• Method 2: In mBlock’s General category, use the Start Tag Recognition block to start the feature.

2. Use tag recognition features.

AprilTag recognition

AI Camera 2.0 automatically marks all detected AprilTags with a white box, assigns them numbers, and retrieves their positions along with 3D data for spatial positioning.

AI Camera 2.0 automatically selects all QR codes in view and displays the decoded information at the bottom of the screen.

AI Camera 2.0 automatically detects barcodes in view and shows the decoded result at the bottom of the screen. It also supports recording via barcode scanners.

Coding block example

Tag recognition blocks

Object recognition

AI Camera 2.0 comes with a built-in large language model (LLM) and identification cards, supporting recognition of 80 types of objects. Additionally, AI Camera 2.0 can recognize the objects selected on the screen. Leveraging the object recognition feature, you can implement projects involving object classification and tracking.

Operation steps

1. Start the object recognition feature:

• Method 1: Tap the Object icon on the AI Camera 2.0 screen.

• Method 2: In mBlock’s General category, use the Start Item Recognition block to start the feature.

2. Use object recognition features.

Object recognition

Select Object Recognition and AI Camera 2.0 will mark the detected object with a green frame and display its name.

Align the white box frame with the target object, press the Learn button or tap the forget icon in the upper right corner. The screen will prompt Confirm to forget?. Tap Confirm to delete the object.

Coding block example

By using AI Camera 2.0 to recognize objects and identify coordinates, you can specify the conditions for determining whether different objects are close. In the following example, AI Camera 2.0 is used to determine whether a person and a mouse are close by acquiring their respective positions.

Speech recognition

AI Camera 2.0 supports Chinese and English speech recognition. With this feature, you can create projects such as a voice assistant.

Operation steps

1. Start the speech recognition feature:

• Method 1: Tap the Speech icon on the AI Camera 2.0 screen.

• Method 2: In mBlock’s General category, use the Start Speech Recognition block to start the feature.

2. Use speech recognition features.

Speech recognition

In English or Chinese mode, speak in the selected language. AI Camera 2.0 will recognize your speech and display the detected text in English or Chinese at the bottom of the screen.

Speech learning

Speech deletion

Coding block example

Character recognition

Operation steps

1. Start the character recognition feature:

• Method 1: Tap the OCR icon on the AI Camera 2.0 screen.

• Method 2: In mBlock’s General category, use the Start OCR block to start the feature.

2. Use the character recognition feature. AI Camera 2.0 highlights detected Chinese or English characters with a green box, and the recognized text will be displayed in the upper-left corner of the box.

Coding block example

Posture recognition

AI Camera 2.0 comes with 12 built-in postures and recognition cards. The posture recognition feature provides the coordinates of each key point, enabling more precise interaction. With this feature, you can create motion-sensing robot projects, similar to the robots in the movie Real Steel.

Operation steps

1. Start the posture recognition feature:

• Method 1: Tap the Posture icon on the AI Camera 2.0 screen.

• Method 2: In mBlock’s General category, use the Start Posture Recognition block to start the feature.

2. Use the posture recognition feature. Select Posture Recognition. AI Camera 2.0 can recognize 12 common postures and display a human skeleton model. You can turn skeleton point detection on or off by tapping the human icon in the upper-right corner.

Coding block example

Path recognition

AI Camera 2.0 supports path recognition based on learned line colors and can obtain the path offset.

Operation steps

1. Start the path recognition feature:

• Method 1: Tap the Line icon on the AI Camera 2.0 screen.

• Method 2: In mBlock’s General category, use the Start Line Recognition block to start the feature.

2. Use line recognition features.

Line learning

Press the Learn button or tap the target position on the screen. AI Camera 2.0 will learn the LAB color threshold at that position, and an arrow in the center of the screen will indicate the forward direction.

Line deletion

Tap Line in the lower right corner to view stored lines. Then tap the forget icon in the upper-right corner to delete a learned line, or overwrite it by learning a new one.

Coding block example

Path recognition blocks

By using CyberPi, AI Camera 2.0, and mBot 2, this example implements line-following control based on path offset. By adjusting wheel speed differences, this example demonstrates the autonomous driving process from line recognition and offset detection to automatic correction.

Taking photos and videos

AI Camera 2.0 also supports taking photos and videos. You can take photos or record videos using the Camera feature and preview or view them in the Photos library.