Computer Vision / Video Analytics

Metropolis Spotlight: Lumeo Simplifies Vision AI Development

Over a billion cameras are deployed in the most important spaces worldwide and these cameras are critical sources of video and data. It is becoming increasingly important to understand how to harness this data to make our spaces and processes more efficient and safer.

Lumeo, an NVIDIA Metropolis partner, provides a ‘no-code’ video analytics platform that enables developers and solution providers to create and deliver custom analytics at the edge or in the cloud in minutes. It accomplishes this through web-based drag-and-drop tools, pre-built analytic building blocks, ready-to-use and custom AI models, and integrations with cloud services.

Accelerated by NVIDIA

Lumeo’s platform builds on NVIDIA pretrained models, NVIDIA TAO Toolkit, and the NVIDIA DeepStream SDK, which pave the fastest path to market for vision AI solutions. You can quickly deploy and scale your custom-built solution on-premises or in the cloud, using any NVIDIA GPU or NVIDIA Jetson edge AI system, and then manage it from the cloud.

For enterprise customers, Lumeo uses Metropolis Showcase to enable them to quickly build solutions and try the AI platform using their own cameras and video files, with nothing but a browser. Metropolis Showcase accelerates development and evaluation by providing customers with access to powerful cloud-hosted GPUs.

“Partnering with NVIDIA is a natural fit as we help empower developers across a wide range of verticals to accelerate their go-to-market,” said Devarshi Shah, CEO and co-founder of Lumeo. “AI is the leading technology for advanced analytics, and we are excited to lead the revolution with NVIDIA by putting businesses in control of their video data and saving them time and money.”

Build vision AI solutions faster

Screenshot of the Lumeo dashboard to create a parking monitoring app.
Figure 1. Easy-to-create AI applications with drag-and-drop builder in Lumeo

Building vision AI solutions presents many challenges, such as managing cameras, processing videos, storing and transferring data, deploying AI models, and using inference results. Lumeo helps alleviate these obstacles with some of its core features.

Using Lumeo, you can manage different inputs like ONVIF cameras, USB cameras, RTSP streams, and video files, all in one place. Deploying pipelines from any video source to the cloud or on-premises devices, standalone or in a Kubernetes cluster, is possible with just a click or API call.

Build video pipelines with the drag-and-drop builder using nodes for recognition, detection, AI models, rules, integrations, media capture, and streaming. It enables enterprises to easily connect their video pipelines to NVIDIA GPUs for inference.

The application workflow involves encoding, decoding, streaming, high-performance processing, AI models, and so on. In the Lumeo web console, you can also review tailored analytics and create custom integrations using REST APIs.

“Lumeo places powerful tools at your fingertips, letting you build your own video analytics to gain actionable insights into business processes and increase the ROI of existing systems without deep technical expertise,” said Shah.

Deployed in real-world settings by integrators and software teams

Video 1. Build a vision AI application in 5 minutes

Lumeo is currently being used to power many use cases, ranging from parking management and people counting to wait-time estimation and license plate recognition. In one instance, a software vendor used its platform and NVIDIA GPUs to help improve parking operations by helping parking management monitor traffic flow, the number of vehicles in an area, vehicle speed, and garage occupancy levels.

The no-code approach also empowers integrators that have previously been limited to deploying existing software. A high-end retail mall needed to keep track of occupancy and people waiting in line. Using the Person Detection, Occupancy Monitor, and Line Crossing models and NVIDIA Jetson modules for edge computing, the mall’s integrator quickly created and deployed a solution that used existing camera and VMS infrastructure to make the retail experience smoother.

Free account

Lumeo recently launched a free tier where you can quickly sign up and get started building powerful vision AI applications in the cloud. Then, deploy the apps to your own devices for production. Try out the platform today by signing up for a free account.

Discuss (0)