Machine learning company SiMa.ai has launched Palette Edgematic, a free visual development environment designed for any organisation to get started with and accelerate ML at the edge. With the new product, the company is set to deliver artificial intelligence (AI) and machine learning (ML) at the edge via a no-code approach to creating, evaluating, and fine-tuning ML applications from anywhere in the world via a secure browser. 

Palette Edgematic enables a “drag and drop,” code-free approach where users can create, build, and deploy their own models and complete computer vision pipelines automatically in minutes versus months while evaluating the performance and power consumption needs of their edge ML application in real time. ML applications evaluated and developed with Palette Edgematic can later be deployed and executed on a SiMa Developer or Production board.  

It is a free visual extension of SiMa.ai’s Palette software platform for end-to-end application development, of which the production (1.0) version was recently released. Powered remotely by SiMa.ai MLSoC, the initial release of Palette Edgematic delivers a GUI-based visual experience to evaluate five optimised pipelines and 40 different models covering object detection, tracking, classification, semantic segmentation, and instance segmentation use cases. 

With Palette Edgematic, anyone with a computer vision pipeline concept can evaluate it directly on an edge device without an intermediate step in the cloud, providing a direct path to implementation at the edge. With Palette Edgematic, a developer will be able to rapidly prototype and evaluate ML pipelines on edge devices in minutes. It gives the facility to use real-time data streams to measure KPIs and then use the Palette Edgematic new visual canvas to rapidly iterate design to improve pipeline performance, eliminating the specialised embedded coding translation for edge ML implementations among others.  

"Only SiMa provides unparalleled innovation for the embedded edge. The release of Palette Edgematic together with our continued dominance in leading the industry in FPS/W performance is a force  – we are setting a new standard in simplifying edge ML once and for all,” said founder and CEO Krishna Rangasayee. “SiMa is laser-focused on removing any friction or obstacles for organisations looking to embed AI and ML into their products and services."

"While machine learning has been widely deployed in the cloud, ML adoption at the edge has been very slow due to a lack of easy-to-use software tools. Customers have to hand-optimise models and the entire end-to-end application to get the necessary performance and accuracy. It typically takes multiple weeks to evaluate and several months to deploy ML applications,” said Gopal Hegde, senior vice president of engineering and operations at SiMa.ai 

“Our Palette Edgematic software provides a no-code visual GUI platform that enables customers to evaluate our MLSoC platform in a few hours. Over time, customers with no ML development experience can develop, fine-tune, deploy, and scale ML applications quickly, democratising ML at the embedded edge,” he added.