Skip to main content

Your first project

This guide builds a face-detection workflow and runs it on a video.

You are done when the preview shows the input frames with the detected faces.

Before you begin​

Make sure you have:

  • access to Clea Studio AI;
  • an input video in your Studio AI workspace.

If you are new to the terms used here, see the Glossary.

1. Create the project​

  1. Open Clea Studio AI.
  2. Create a new project and name it Face Detection on Camera.
  3. Open the project editor.

The left panel groups the blocks into Input, Model, and Output.

2. Build the workflow​

  1. Place these blocks on the canvas: Video Input, Face Detection HMI, and Image logger.
  2. Connect the image output of Video Input to the input of Face Detection HMI.
  3. Connect the detection output of Face Detection HMI to Image logger.

The result is:

Video Input > Face Detection HMI > Image logger

3. Configure the input​

Select Video Input and set:

Source: your video file
Drop policy: target-fps
Drop policy value: 15
Frame width: 320
Frame height: 240

Save the project.

4. Run the preview​

Start the preview from the project editor and watch the output of Image logger.

Check that:

  • the input video is read correctly;
  • the face detector receives frames;
  • detections are visible in the output;
  • the project remains in a running state.

Next steps​

Take the project to a device:

To connect a device first, see Get Started with Clea 2.0.