AWS Panorama now supports the ability to bring your own machine learning libraries

You can now run models and algorithms from TensorFlow, PyTorch, and OpenCV directly on the AWS Panorama Appliance GPU for low-latency use cases, such as object tracking, traffic analysis, and worker safety. AWS Panorama also supports models optimized with Nvidia TensorRT or any machine learning library with Nvidia CUDA support. Nvidia TensorRT can double the number of camera streams that can be processed simultaneously on the AWS Panorama Appliance for a given use case.

AWS Panorama is a machine learning (ML) appliance and software development kit (SDK) that brings computer vision (CV) to on-premises internet protocol (IP) cameras. With AWS Panorama, companies can use compute power at the edge (without streaming video to the cloud) to improve their operations, by automating visual inspection tasks like evaluating manufacturing quality, finding bottlenecks in industrial processes, and assessing worker safety within their facilities.

You can start building applications for AWS Panorama by checking out our sample apps, which covers how to run TensorFlow, PyTorch, ONNX and Nvidia TensorRT directly on the GPU.

The AWS Panorama service is available in the US East (N. Virginia), US West (Oregon), Canada (Central), Europe (Ireland), and Asia Pacific (Sydney and Singapore) regions. The AWS Panorama Appliance is available for purchase on Amazon.com, Amazon Business and AWS Elemental in the United States, Canada, Mexico, Australia, New Zealand, Singapore, New Zealand, Malaysia, United Kingdom, and the European Union.

To learn more about AWS Panorama, visit the AWS Panorama product page.