Neurotechnology Announces the Release of SentiVeillance Cluster, Ready-to-Use Software for Surveillance Systems with Clustered Architecture Implementation for Smart-city-scale Projects

Neurotechnology, a provider of deep learning-based solutions and high-precision biometric identification technologies, announced the release of SentiVeillance Cluster for real-time biometric face identification, tracking of people and vehicles and automatic vehicle license plate recognition. The solution has been designed for expansive surveillance systems supporting continuous video streams from multiple servers. It provides operators with structured and sorted live data, enabling faster decision-making in a variety of scenarios, including law enforcement, security and smart city monitoring applications.

“We are thrilled to present SentiVeillance Cluster, as it will open up a new range of opportunities for our clients,” said Vytautas Pranckėnas, SentiVeillance product lead at Neurotechnology. “Equipped with scalable architecture and award-winning biometric algorithms, this solution offers flexible applications and reliable results. We see immense potential for this technology to be used in various applications – from small surveillance setups to entire smart city monitoring systems.”

Recommended AI: Microsoft 365 Security Features Protect Business Data from Evolving Threats

The SentiVeillance Cluster software is designed to analyze numerous video streams at once and combine multiple servers into a cluster network. The system’s key features include:

• Biometric person identification and tracking
• Pedestrian as well as vehicle tracking and classification
• Integration with video management system Milestone VMS
• Real-time custom watchlist check with automatic event triggering and logging
• Multiple video streams that can be analyzed on a cluster of machines
• Multiple GPU support

SentiVeillance Modalities:

The SentiVeillance Cluster software supports biometric face recognition, a vehicle-human modality and automated license plate recognition (ALPR) modalities for surveillance systems.

• Biometric face recognition
  In addition to continuous face tracking and person recognition, the face algorithm features gender classification, age determination, and attribute detection capabilities. Watch lists can be generated using face image data and additional metadata (e.g., biographic fields, custom fields) to populate watch lists quickly. No additional enrollment is required for face coverings (e.g. masks).
  In access control applications, face recognition ensures a high degree of confidence that only authorized people can get access to premises both in the government and private sectors. SentiVeillance’s monitoring for safety and security decreases the likelihood of theft, abuse, gang violence and other illegal activities.

Recommended AI: Consider Your DOOH Buying Methods Wisely: Direct Sales vs. Programmatic Buying

The algorithm’s accuracy and performance have been tested in the most reliable series of large-scale, independent evaluations for face recognition algorithms at NIST Face Recognition Vendor Test (FRVT).

• Vehicle-human modality
  Vehicle and pedestrian detection can be performed on moving or static objects, allowing system users to perform operations directly from live video streams. The algorithm can distinguish between vehicles such as bicycles, trucks, buses, cars and pedestrians. Additionally, estimates of color, vehicle brand information and movement vector can be made.
  In electronic toll collection systems, SentiVeillance algorithms can be used for quick and accurate vehicle recognition to automatically collect the usage fee or toll charged to the vehicle’s owner for using toll roads, tunnels, etc.

• Automated license plate recognition (ALPR) modality
  The ALPR algorithm offers reliable detection, even in difficult recognition scenarios, allowing high tolerance to the camera position and license plate orientation. Using the ALPR and VH modalities together, it is possible to verify if a license plate is on the correct type of vehicle and alert authorities if it is not.
  The system can simultaneously analyze multiple traffic camera videos and read numerous moving vehicles’ license plates in real time. This feature provides parking lot security by using an automated surveillance system to reduce and prevent theft and damage. Surveillance systems with the ALPR modality installed at parking lot entrances guarantee access control for registered vehicles.

Parking lot surveillance application

SentiVeillance customer uniPark implements parking solutions throughout the Baltics and Poland, with more than 160 projects in Lithuania, including airports, shopping and business centers and city center parking zones. uniPark worked together with the SentiVeillance team to create car park monitoring, without conventional barriers or registration terminals, based on automated license plate recognition and video feed from regular surveillance cameras.

“We’ve been using the SentiVeillance solution in uniPark parking lots for over a year, and we are absolutely impressed with the benefits of this technology,” said Justinas Doviltis, director of IT at uniPark. “The ability to register cars from different camera angles and in totally uncontrolled scenarios without barriers or closely placed cameras gives us new opportunities to provide flexible and cost-effective solutions to the market.”

Recommended AI: Understanding the Role of AI in Gaming

[To share your insights with us, please write to sghosh@martechseries.com]