Video Analytics Platform VisualCortex is the New CV Kid in Town

A new video analytics platform, VisualCortex, has been launched globally. The company says its enterprise-grade platform is capable of facilitating any real-time or historical video analytics use cases, unlike single-use-case solutions that are less scalable or adaptable to an organization’s existing data stack.

VisualCortex says its mission is to “enable any organization to become a vision aware enterprise, allowing them to solve commercially valuable challenges with video-based insights at scale.”

Like many facets of AI technology, those embarking on computer vision projects often encounter barriers to entry. This is especially true for larger projects as there are often high costs involved for software and computing resources, data scientists to develop and deploy CV models, and high quality training data for modeling.

VisualCortex was designed to work with streaming and offline footage alike, and the platform applies machine learning models to video content, “enabling organizations to produce analyzable data streams about defined objects and actions. Insights from that data can then be embedded straight into customers’ traditional data infrastructure.”

VisualCortex is backed by Tony Nicol, the co-founder of Australian data cloud company Servian. Nicol founded Servian in 2008 and was CEO until June 2021 when the company was acquired by U.S. professional services firm Cognizant. It was during that time that Nicol gained experience building large scale enterprise data platforms, enabling him to build what he says is a flexible video intelligence solution that provides stability, security, and governance capabilities while delivering tangible value.

An example of a CV model for monitoring vehicle attributes. Source: VisualCortex

“For the first time, we’re making video data truly actionable throughout the enterprise,” said Nicol. “Up until now, computer vision technology has struggled to make commercial sense and generate impactful business value. They’ve also been prohibitively hard-to-use and expensive in terms of cost and time-to-value, hampering the ability to be harnessed by anyone other than machine learning experts.”

For those who do not wish to develop their own CV models, the VisualCortex platform also features a model store where customers can find ready-made models designed, developed, and trained by the VisualCortex team, as well as some third party models contributed by the company’s clients and partners. There are vehicle and personal attribute models, models for object velocity and motion analysis, and many other examples.

The company says its VisualCortex platform was designed to be deployed in a variety of ways. The company offers a fully managed proprietary cloud option, or users can leverage it on private clouds, public clouds, on-premises, or at the edge. They can also choose a hybrid approach. It was designed to be accessible to anyone, including ML and data engineers, business analysts, solutions architects, C-Suite executives, and others. Additionally, organizations can use their existing video infrastructure and software stack. They can also choose to use the platform’s own analytics functions, or it can be integrated with other data analytics platforms.

This graphic shows the platform’s integrations and capabilities. Click to enlarge. Source: VisualCortex

VisualCortex CEO and Co-Founder Patrick Elliott is confident of the platform’s flexibility for all types of industries, use cases, and users.

“VisualCortex is making video intelligence accessible and valuable for all video-rich industries and business functions,” said Elliott. “Our Video Intelligence Platform provides the AI smarts, governance and stability. Clients just need to bring their standard infrastructure, commodity hardware and video – from any stream, camera or repository. With VisualCortex, you can run multiple machine learning models, for any number of video sources, across your existing cameras and video feeds.

“We’re enabling anyone to quickly connect machine learning models with a production-ready cloud-based environment, transforming video assets into analyzable and actionable streams of data at scale. VisualCortex empowers technical users, like software developers and data scientists, as well as business users – from marketing executives and business analysts, to operations teams for OHS, security, car parks and more.”

Related Items:

How Data-Centric AI Bolsters Deep Learning for the Small-Data Masses

Computer Vision Platform Datagen Raises $50M Series B

How Coke Bottlers Save Millions with AI