March 22, 2017 -
Neurotechnology has released SentiVeillance 6.0 software development kit (SDK), which provides improved facial recognition using up to 10 surveillance, security and public safety cameras on a single computer, making it suitable for a wide range of surveillance applications.
According to a company statement, the new version uses deep neural-network-based facial detection and recognition algorithms to improve accuracy, and it utilizes a Graphing Processing Unit (GPU) for enhanced speed. In addition, users can more quickly and easily adjust the tradeoff between speed and accuracy as needed for different applications.
“We developed SentiVeillance 6.0 as a self-adapting system based on deep neural networks that were trained on a larger quantity of data,” said Ignas Namajunas, surveillance technologies research lead for Neurotechnology. “This ensures better generalization for a variety of conditions. Additionally, by making use of the GPU processing capabilities, we were able to improve the processing speed significantly.”
As with previous versions, the new SentiVeillance provides: real-time moving object detection; tracking and classification for pedestrians, vehicles and other predefined object classes based on size and speed of movement, and; area control that triggers “events” when people or objects enter, leave or stay in restricted areas.