/ Real-Time AI

Easy-to-Use Real-Time AI Framework

The team at KickView is very excited to announce the release of our easy-to-use, real-time AI video and multi-sensor processing framework, called kvSonata!

Advances in machine learning, computer vision, and multi-sensor signal processing have opened new opportunities across nearly every industry. The potential benefits of using AI-based technologies are currently unbounded. However, in too many situations, product developers and researchers are required to spend valuable time and resources coding up low-level infrastructure in order to create real-time sensor processing solutions.  This is even more difficult when access to AI talent is not possible. Additionally, platforms and SDKs offered by large hardware companies are aimed at encouraging product developers to use specific proprietary hardware. These platforms and SDKs are complex to develop in - requiring expert knowledge of a specific low-level programming language and often requiring many days or weeks of effort just to install and configure basic examples.

To make developing real-time AI solutions easy and more efficient, KickView has created the kvSonata framework for real-time video and multi-sensor AI processing.  KickView has been succesfully using kvSonata for almost two years to create solutions for industry leading customers. Now, we are excited for you to experience kvSonata!

kvSonata allows you to use both modern AI techniques and traditional processing algorithms from computer vision, signal processing, and machine learning. You can even create custom processing modules in Python, C++ or Golang! Also, you are not required to use a single hardware vendor solution. Go ahead, use a CPU, VPU, or a GPU! If you need help creating a custom solution, we can provide expertise and development for your project.

Let's talk about our favorite kvSonata features:

  • kvSonata takes care of the data pipelining and processing module structure for real-time data processing
  • kvSonata let's you develop and add code in multiple languages (e.g., Python, C++, Golang)
  • Did we mention it is easy-to-use?
  • kvSonata provides container-based, scalable architecture for real-time processing
  • Many use-case examples are provided to get you started creating solutions
  • Compatible with multiple hardware targets - GPU, CPU, VPU, and more to come
  • Supports many machine learning and deep learning frameworks
  • Provides time-synchronization of multiple sensor inputs
  • Lets you run multiple independent user-defined processing workflows across multiple sensors
  • Quickly prototype, develop, test and deploy in one framework
  • kvSonata will save you time...lots of time!

For example, see the image below, a tiled combination of four different precessing modules.

The video tile shown above, which includes 6 different modules (stream capture, TensorFlow object detection, edge detection, blur filter, tile, and a display module), can be defined with the simple YAML file shown below. If you want to add more video streams, just configure more inputs. If you want to add another module to the pipeline, all you have to do is add another section to the YAML file and push play!

(Above Figure) YAML workflow definition of the video processing tiler shown above. Notice that modules written in Python, C++, and Go are all run together in the same workflow.

The kvSonata platform is perfect for real-time video analytics solutions. However, we should also let you in on a secret. You can also use it to make efficient use of data from other sensors such acoustic, radio frequency (RF), IoT, and more! Once configured, a kvSonata workflow can be created to process, analyze, fuse and display important information, all in an easy-to-use, completely configurable manner.

Who can use kvSonata?

Do you work with any kind of sensor data, such as video, acoustic, or RF signals? Getting started with kvSonata is quick and easy. In fact, we believe it is the most developer/user friendly framework for real-time video analytics available. It is also perfectly accessible for non-experts. kvSonata gets you working on solutions quickly and lets you focus on how you want to configure, process, analyze, and display results.

kvSonata is intended for use by:
• Product developers
• Edge computing product designers
• Engineers working in AI, computer vision, IoT, or robotics applications
• Researchers
• Anyone who wants to process data in real-time

Because of its easy-to-use plug-and-play architecture, fully-customizable module templates, multi-programming language support, and containerized environment, kvSonata is the go-to solution for any of the following use cases:

  • Real-time intelligent video analytics
  • Real-time computer vision processing
  • IoT application development
  • Machine learning algorithm prototyping and testing for streaming sensor data
  • Real-time sensor fusion
  • Fast prototyping of real-time AI solutions
  • Deep learning projects
  • And more…
(Above Video) Real-time AI parking application created with kvSonata framework.

At KickView, we have been using kvSonata for developing word-class real-time AI solutions for video and multi-sensor applications. We have created solutions for smart cities, retail, traffic, parking, security, and government applications. We want to help you avoid the complexities and cost often associated with creating real-time AI solutions. If you think we can help you with a solution, don't hesitate to contact us at https://kickview.com/contact/.

In the meantime, if you are interested in being notified for kvSonata early release, please submit an inquiry at https://kickview.com/contact/, and stay tuned for the upcoming series of blog posts about how to use and benefit from kvSonata.

For more information, see https://kickview.com/kvSonata/



KickView provides real-time AI solutions for extracting actionable information from sensor data at the edge.

Read More