You Don’t Have to Go It Alone

NVIDIA's Metropolis Microservices for Jetson accelerate edge AI development by providing a simple API to handle complex vision AI tasks.

Nick Bild
6 months agoMachine Learning & AI
NVIDIA's Metropolis Microservices for Jetson accelerate edge AI tool development (📷: NVIDIA)

The pace of change in artificial intelligence (AI) is accelerating at an unprecedented rate, ushering in transformative advancements that seemingly impact every aspect of our technological landscape. Between autonomous vehicles, medical diagnostics, and agricultural monitoring, the innovations that we have seen in the past decade can be enough to make your head spin.

For the technologically-inclined among us, this rapid pace of change can make us feel like we are being left behind. After all, there is a sense that big things are happening, and yet most of us have done no more than stumble our way through a cat versus dog classification tutorial, then go no further because the complexities quickly grow too great.

For those whose day job does not directly intersect with developing AI applications, keeping up to date with the latest advances — not to mention the techniques that power those advances — may not be realistic. So what is a person supposed to do if they have a bunch of really good ideas for AI-powered applications, but no idea how to make them a reality? Metropolis Microservices for Jetson, which was just announced by NVIDIA, may be a good resource in these situations. They have taken one of the most complex of all AI applications — vision-based tools — and created a simple API-driven edge AI development workflow. Metropolis provides a number of microservices that handle common tasks required of vision AI applications, while hiding the complex details.

By utilizing this toolkit, a developer can focus on an idea rather than the complexities associated with implementing it. In this way, the microservice building blocks can save a developer a lot of time, and also enable them to put powerful tools at their disposal that would otherwise be out of reach. These microservices are accessible via an API that is based on a standard pattern used in cloud-native architectures. At present, fifteen different microservices are available for video storage and management, prebuilt AI perception pipelines, tracking algorithms, system monitoring, IoT services for secure edge-to-cloud connectivity, and more.

Using the NVIDIA Jetson hardware platform with Metropolis Microservices, production-ready edge AI applications can be built in a fraction of the time required in a traditional development cycle. By leveraging prebuilt microservices to ease the creation of AI models, optimized inference pipelines, security procedures, cloud connectivity, and so on, the usual long and costly development cycles can be short-circuited. For large, complex systems, these tools can shave off months, or even years, of development time.

A pair of sample applications built with the Metropolis for Jetson platform have been provided by NVIDIA to assist in teaching the important concepts and to help developers get their own ideas off the ground quickly. These applications are an AI-enabled network video recorder and a generative AI application with zero-shot detection capabilities. Demonstrations of some of the most important microservices, and how they can be integrated with one another, are contained in these sample applications.

More details about NVIDIA Metropolis Microservices for Jetson are available in the press release. A how-to guide has also been published to help new users in getting started with the platform.

Nick Bild
R&D, creativity, and building the next big thing you never knew you wanted are my specialties.
Latest articles
Sponsored articles
Related articles
Latest articles
Read more
Related articles