Real-Time Intelligent Video Analytics For Your Applications

If your mission is to deploy effective edge processing solutions while having direct access to expert machine learning resources and support, then this blog post is for you.

Real-Time Intelligent Video Analytics For Your Applications

It makes sense that everyone is interested in integrating artificial intelligence (AI) into their applications and products. Not only is AI very interesting and popular, but it opens new avenues for innovation and increased performance. However, it can be a challenge to effectively implement AI, or machine learning (ML), into real-word sensor systems. To do this, you often need to take models that were trained in the lab, with currated data, and deploy them into a system ingesting 24/7 streaming sensor data. Not only will you need to understand how to train these models, but you will need to know how to architect a data pipelining system that can support streaming data from multiple sensor inputs.

KickView can save you time and money by utilizing kvSonata - the world's best real-time AI processing framework for intelligent video and sensor applications. We excel at providing bespoke solutions and custom applications that automate customer processes - making them efficient, intelligent and fast!

If your mission is to deploy effective edge processing solutions while having direct access to world-class machine learning resources and support, then KickView's kvSonata is for you.

Figure 1. Real-time detection and geospatial tracking from moving camera sensor using kvSonata. Drone camera view (left), Google map with fused object tracking.


When dealing with low-latency applications, utilizing an easy-to-use framework can be the difference between deploying solutions quickly, or spending too much time and money on complex and fragile infrastructure. KickView has created the kvSonata framework to excel at real-time intelligent video and sensor processing at the edge. It enables edge solutions to be configured and deployed extremely fast. In addition, those applications can be easily reconfigured, as new sensors are added or when system requirements change.

Figure 2. KickView's kvSonata framework for easy deployment of real-time intelligent video and sensor analytics at the edge.

Typically, getting value from AI/ML requires multiple development steps, tools and technologies. Organizations and development teams are slowed when they have to develop a data pipelining architecture and train and deploy custom algorithms and models. It is better to manage and apply machine learning in a real-time production environment using a single adaptable framework. That is what kvSonata provides.

Figure 3. kvSonata workflows enable quick prototypes, tests, and deployments of intelligent video processing solutions that leverage GPU, CPU or VPU (or mixed) resources. Quickly try and test computer vision and ML algorithms for your applications.

KickView's real-time AI framework kvSonata was created using a containerized design, so it will be transportable to your compute platform, be it an edge computer, device, server, or cloud resource. It also allows you to leverage a CPU, GPU, VPU, or combination of them for optimized AI processing. Prototype, test and deploy in a single framework, fast!  

Figure 4. Smart parking application created with the kvSonata framework using real-time intelligent video processing.

Combining your video data with real-time AI to generate intelligent analytics (i.e., actionable intelligence) can greatly benefit your organization. From using video for preventing loss, to leveraging your video sensors for identifying incidents, real-time intelligent video analytics makes for better and faster decisions. Learn more about how to get started using kvSonata in your applications by contacting us for a demonstration and access to evaluation software.

Our Enterprise Package inludes full access to the KickView library of machine learning models and custom training services, along with concierge engineering support. This premium package allows you to create custom applications and solutions with both your data sets and KickView's customized models -- no need for data science expertise. The kvSonata framework allows for fast edge applications to be configured for security, surveillance, smart cities, intelligent traffic control, parking, retail, automated inspection and verification, test and measurement, spatial tracking, and more.