Run computer vision models inside ACAP applications
The Axis Camera Application Platform (ACAP) enables you to build and deploy applications that run deep learning models directly on Axis cameras. By integrating a trained model into an ACAP application, you can perform real-time video analysis on the device, reducing reliance on cloud-based processing.
Create and prepare your model
For guidance on training, converting, and optimizing a deep learning model for Axis devices, see the Computer vision documentation.
Integrate a model with an ACAP application
To use your model inside an ACAP application, use the Video capture API (VDO) to get frames from the camera and the Machine Learning API (Larod) to run DLPU-accelerated inference on your model. Explore the following examples in the ACAP Native SDK examples repository to get more inspiration about how this can be implemented: