AI-Driven Video Understanding Technology

Request a demo


viisights’ innovative behavioral recognition video understanding technology is based on a unique implementation of deep neural networks. These AI-driven networks are capable of analyzing and deducting high-level concepts derived from video content.

viisights is developing unique video understanding technology that is capable to understand video content in a very similar way to humans viisights technology uses state-of-the art
computer vision artificial intelligence which is using a unique architecture that detects objects and their attributes, and on top of that focuses in detecting their behavior in non-sterile
environment, like a city street. This architecture is capable to detect dynamic scenarios like events (e.g., people fighting), actions (e.g., person holding a rifle in threatening position) and
scenes (e.g., brawl in a crowd).

On top of that technology viisights is building products for the markets of video security & safety which offer significantly more relevant, accurate and advanced features than common analytics
systems which are mostly based on static classification of objects. This approach manifests itself by. viisights capability to detect how many people are in an area and on top of it detects what
the people are doing and how they interact with each other.

viisights technology is based on proprietary machine learning algorithm trained on both images and video clips. The trained inference engine uses temporal analysis and holistic view of
complex scenes with multiple objects interacting between themselves. All these is done in real-time in a highly efficient way, thus having low hardware footprint and low cost per live stream.

The integration of multiple deep neural networks and the holistic analysis of multiple dimensions of video understanding, creates a significant challenge in terms of throughput and scalability processing.

viisights overcomes these challenges in a two-folded approach:

Using NVIDIA GPU processors which provide the demanding processing power required by the system.

Incorporating a unique system architecture that significantly shortens the processing time of each analysis aspect, thus allowing the system to complete the holistic analysis in near real time.



Our intelligent behavioral recognition video understanding technology utilizes a unique orchestrated architecture that enables the effective implementation of advanced deep neural networks. This architecture supports multiple holistic views by using:

  • Multi-scale image analysis
  • Time-aware analysis
  • Smart integration between object detection and the tracking mechanism
  • Innovative object detection structure that accelerates performance and increases accuracy


viisights behavioral recognition video understanding technology leverages the classical deep learning model, as depicted in the following diagram, while implementing unique training tools and methodologies to reduce the amount of data required for training


viisights uses NVIDIA GPU in both the training process and in the production system. In the training system it uses the Tesla P100 GPU and in the production, it uses several types of GPU depends on feature set configuration and the required workload. By such choice of hardware, viisights combines the power of the advanced GPU with affordable cost in the production environment.