Using Sense, you'll be able to:

• Add compute nodes and cameras

• Create training clips from camera image data

• Label training data using HyperLabel

• Train ML models with your data sets

• Deploy models to your camera, set of cameras, or to an entire node

• View the detections that result from your model

• Monitor your model's performance

1. Add a node

Sense's edge processing software is called a node. It handles the computational tasks such as ingestion of video data for model creation, the deployment of trained ML models, and gathering detections from the data.

Learn how to add a node and get started with edge processing.

2. Add your cameras

Sense lets you connect devices, such as cameras, which can be used to train and deploy your machine learning models to detect objects. Connect your camera to Sense to give your node access to the video stream.

Read more about how to add your camera and assign it to your node.

3. Create your model

Create your model to begin adding training data from your video data, then label frames using HyperLabel.

See how easy it is to create a model.

4. Create training data and begin labeling with HyperLabel

You can create training clips from your video data, then label the frames using HyperLabel.

Find out how to create training data sets and check out our HyperLabel guides to start labeling.

5. Train and deploy your ML model

It's never been easier to train your model with your labeled data sets and deploy the latest versions to your camera.

See our guide on Training and Deployment for more information.

6. Monitor your model's performance

Use the "Monitor" tab to explore your detections in real-time from each of your models. You can view charts to see trends in the data and drill into the timeframes for a closer examination.

​