AI parameters

In the AI Parameters tab of the Camera settings section, you are able to change the AI resolution and AI model used by that particular camera. This is used to manage the costs of using AI for your cameras. You have the options

Resolution:

• HD (High Definition)

• FHD (Full High Definition)

Model:

• Lite

• Pro

• Advanced

Resolution

The resolution determines the quality of image the AI model will use while performing it’s functions. HD images have a slightly lower resolution than FHD and will result in less crisp images to perform queries on.

Model

The model determines which LLM (Large Language Model) will be used to perform queries on. Some models are more sophisticated than others, hence the Advanced, Pro, and Lite naming convention. The specific names of these models are:

Credits

Credits are required for making calls to AI. One credit can be used for the following number of calls:

When you change between models and resolutions, we convert your AI Credits in the backend so credit usage increments much slower for light models and lower resolutions.

Models and use cases

Objects and scene description

Usually Lite model will work well for the use case that requires a scene and object description. For example:

• Are there more than 3 people near the elevator?

• Is there a man in a red shirt?

Scene analyses

If you require more detailed scene analyses then Pro model is recommended. Examples of prompts:

• Are there many shopping carts near the self checkout?

• Is there a child coming back home?

Context analyses and anomaly detection

For deep context analyses and anomalies detection we recommend using Advanced model. Here are some examples of prompts:

• Is there any security concern?

• Is there a safety hazard?