PROIECT: VISINT

The solution studies human behaviour, based on a neural network architecture that classifies human gestures and emotions based on body movements, facial features and variations in physiological parameters. In the first instance, the system captures biometric data correlated with certain behavioural tendencies.

Subsequently, by developing a complex of convolutional neural networks (CNNs), an automated behavioural analysis tool is implemented, exploiting different learning techniques.

The system is effective in identifying emotional states in real time for behavioural assessment of critical infrastructure personnel.