ML for Embedded Linux: Edge Impulse Announces NVIDIA Jetson Nano Support

With the addition of the NVIDIA Jetson Nano, you can now build ML models for the entire embedded spectrum using Edge Impulse.

We're excited to announce our official support for the NVIDIA Jetson Nano!

Now, users of Edge Impulse can leverage the power of the Jetson Nano for their embedded machine learning applications that demand higher performance, alongside the industry's leading embedded ML platform that offers:

  • The easiest-to-use embedded machine learning pipeline for deploying audio, image classification, and object detection applications at the edge with zero dependencies on the cloud
  • Streamlined acquisition of critical environmental sensor data, previously discarded or only sent to the cloud, for empowering sensor fusion at the edge

We’ve brought the same great user experience our developers are already familiar with into the Jetson Nano domain, with a refreshed set of tools and capabilities that makes deploying embedded machine learning models on the Jetson Nano very… speedy.

In addition, we are also thrilled to launch support for true object detection as part of our computer vision ML pipeline! Plug in a standard USB web camera into one of the available USB slots on the Jetson Nano, and harness the raw power of higher performance compute and more sophisticated frameworks and libraries to facilitate computer vision applications such as OpenCV.

True object detection with an easy-to-use workflow in Edge Impulse
Digits recognition with real-time inferencing on the Jetson Nano
Banana ripeness classification using live feed from Jetson Nano

For audio applications, plug a standard USB microphone into one of the available USB slots on the Jetson Nano. For sensor fusion, the 40-pin GPIO header on the Jetson Nano can be employed to connect to your favorite sensors as well.

Audio classification with real-time inferencing on the Jetson Nano

The best way to get started is by going through our Jetson Nano guide and experiencing the enhanced user workflow for Linux. Then, easily train an object detection model with the help of our tutorial.

SDKs for Python, Node.js, Go, and C++ are provided so you can easily build your own custom apps for inferencing. Furthermore, users can export their model along with our SDK to a TensorRT library for easy inclusion in their apps.

Export your model to a TensorRT library for inclusion in your app

Additionally, here is an example using our Node.js Linux SDK that sends a text message using Twilio if a person and elephant are seen in the same frame.

We’d love to hear from you on our forum about what you think and can’t wait to see how you plan on unleashing the combined power of Edge Impulse and Jetson Nano in your embedded machine learning applications!

artiebeavis

Hackster CMO.

Latest Articles