+ Real-time analysis of eye gaze behaviour
+ Best performing machine learning/object recognition
+ Highlights threat areas instantly, with higher accuracy
+ Easy and fast post-detection handling of baggage
+ No intrusion of primary inspection process
+ Monitoring ‘alertness’ of viewer
Using automated analysis of a person’s eye movements, we are able to accurately identify, within seconds, whether a person is seeing a threat – or not – in a scanned image of airport hand baggage.
Based on the machine’s gaze analysis, a first and fast selection of relevant objects is made. This selection is fused with the machine’s ‘view’ of the image (i.e., object recognition using convoluted neural networks), and then it is determined if the baggage contains a threat or not. Application of such support to security agents is expected to result in considerable speedups, fewer false alarms and less misses.
JOA Scanning Technology was founded in 2017 as a spin-out from the Netherlands Organisation for Applied Scientific Research (TNO). The methodology was initially developed under TNO’s Perceptual and Cognitive Systems Department and Geological Modelling Department. JOA Ventures with its Tech Fund JOA Ventures Seed Capital I, is participating investor.
JOA Scanning Technology’s approach is patent pending (Van Maanen & Brouwer, 2018).
Peter-Paul van Maanen
Chris te Stroet
If you have Artificial Intelligence / Computer Science background and you are familiar with experimental data – great! We offer you a challenging and exciting intern/graduation program position in our growing team. You will be working on new machine learning methods to further improve the classification (i.e. search/pass) of 2D or 3D X-ray images of hand baggage at airports.
If you are interested in joining our developers’ team, please send your application letter and CV to firstname.lastname@example.org