FAQ
Details about the the Zensors platform and deployment mechanics.
Yes; there is no limit to the number of different sensors that can be run on a single camera feed. Zensors is designed to use a single camera to replace multiple special-purpose sensors.
Zensors dynamically adjusts upload based on your connection bandwidth. Zensors does not store videos in the cloud. Zensors will also use different upload rates based on the application.
Zensors is designed as a real-time platform and collects data over time for attached cameras, enabling you to look up statistical data in the past. Zensors is not designed for archiving, indexing and searching of historical video footage.
Zensors has a deep learning model that uses a multitude of factors to estimate wait time. These typically comprise of the number of agents visible on staff, historic data on line length, and rate of flow. The Zensors AI does not require you to tune anything and will automatically learn to improve its estimate within the first few weeks of being deployed.
Zensors has a state-of-the-art face-blurring system that can be enabled for your account. This is an account-level setting and will automatically apply to all of your camera feeds. Please contact us to learn more about this capability.
The Zensors AI engine is constantly learning from new images and camera feeds. The Zensors engine randomly selects images from your deployment to retrain the AI model and provide more accuracy on your particular sensors over time. This can depend on the frequency of the event or object that you are looking to identify. Typically, for an object that Zensors has not seen before, it will take from one to three weeks to train a new model, but you will still receive verified high-quality results as soon as you set up the sensor.
Zensors is currently working on integrating with select camera manufacturers for edge-compute capabilities. Please contact us to learn more about possible on-premises solutions.
Zensors provides the capability to aggregate and join results from multiple cameras. Please ask us for a platform demo to see the possible ways of doing this.
Zensors can work with most network-based cameras. We recommend HD cameras or better for best results. We also have a tool to estimate the field-of-view of a camera to help you arrange your cameras for optimal coverage.
We recommend using our ZOPI system, which acts as plug-and-play gateway between your CCTV network and the Zensors platform. It sits inside your network and pushes encrypted images to the platform so that you don't need to compromise your security by creating a hole into your network. In cases when this is not available, we also support several other protocols, such as HTTP uploading and FTP.
In most situations, Zensors data is real-time and available within 30 seconds from us receiving the image.
Zensors is an active learning platform that has an automated system for monitoring performance and accuracy. The ML engine will periodically perform a test to ensure that your deployment is running at human-level accuracy and work to retrain and improve if it deviates.
In addition to our web application, we also provide an API that can be used to connect our data to external tools and applications. We also have support for CSV export from the web application.