Robovision AI (RVAI) is available as a pre-integrated, end-to-end solution for a number of computer vision applications, or as a powerful AI platform to build the applications you need. The RVAI platform features a rich set of interconnected modules that allows easy integration and deployment of deep learning algorithms into a generic process flow. The modules, which are accessible through an intuitive GUI, include advanced deep learning models, pre-designed computer vision pipelines, training modules and monitoring tools.

RVAI aims to maximize flexibility and accessibility. It features a rich SDK for data scientists and engineers to integrate their applications in a scalable fashion. This way, production operators can leverage state-of-the-art AI applications without having to write a single line of code. With the dashboard application, it is very easy to make applications accessible to operators: create a simple GUI by dragging and dropping widgets on a canvas. 

The quality and consistency of AI-powered computer vision applications much depends on the labeling. RVAI was designed to enable labeling for large volumes of data with the highest efficiency and accuracy. An advanced annotation tool allows both predictive labeling and user management for large labeling crowds. As such, there is no limit to scaling your deep learning computer vision applications. RVAI supports labeling for the widest variety of image formats, each of which require very different technology. For this purpose, it features four labeler modules:

• 2D: for RGB and hyperspectral

• 2.5D: combination of 2D and 3D with mapped/interpolated labeling

• 3D point clouds

• 3D density (medical e.g. DICOM data)

RVAI supports cloud, edge  & hybrid deployments, to scale workflows and applications globally and enable collaboration by remote scientists and users. Applications can be shared with other users or organizations through the RVAI Store, which can be set up as a collaborative environment for other departments or a revenue generator for external customers.